Bat sounds

Many textbooks will tell you that you can hear from 20 Hz to20 kHz. Don’t believe them: if you are over twenty, you are probably deaf to sounds above 17 kHz. The high-frequency limit of our hearing declines so significantly and so predictably with age that a youth-repelling generator of higher-frequency sound called the Mosquito has been used by irate shopkeepers since 2009, as only teenage ears (and those of children and babies) are still capable of detecting the annoying tones the device produces. However, it is probably fair to say that we who cannot hear anything about17 kHz are not inconvenienced by the fact, showing that it is of little evolutionary benefit to hear such high notes, otherwise we would have developed more robust hearing systems.

Nevertheless, we are missing out—not because the ultrasonic soundscape is an especially rich one, but because we cannot exploit some handy physical properties of sounds that are negligible at those frequencies that we can hear. The bat, however, is not so handicapped.

The sophistication with which bats employ ultrasound is astounding. In pitch darkness, an Egyptian fruit bat (Rousettus aegyptiacus) with an 80 cm wingspan can easily ffy between vertical rods just 53 cm apart without touching them. To achieve such feats, and to hunt, bats use echolocation—generating ultrasounds and timing the delays until they hear the echoes—which inform them of the distances of nearby objects.

Sightless humans approximate this, judging distances to walls from echo delays, but what they end up with is not a vision of the world. It cannot be, at the wavelengths they can hear: imagine a jetty standing in a calm lake. A light breeze blows, forming ripples about a centimetre in wavelength, which bounce ofl the jetty’s columns in circular patterns (Figure 21). Later, a gale sets in, making much longer waves, which roll past the columns as if they’re not there. For waves of sound, the same principle applies: they are only aflected by (and hence can only detect) obstacles larger than they are.

So, if a bat tried to echo-locate a moth using an audible tone—middle C, let’s say, which has a wavelength of 132 cm—it would only succeed if the moth were well over a metre across.To detect a 1 cm moth, sound waves of at least 33 kHz are required. In fact, bats typically produce sounds of 80 kHz (smaller species produce higher frequencies, and the full range is about 40 kHz to 120 kHz). Such high frequencies allow them to deal not only with isolated objects, but with complicated three-dimensional distributions of twigs, leaves, and insects, many in relative motion to each other—and all in relative motion to the bat.

But it would do the bat no good to generate a continuous ultrasonic whine: the result would be a bewildering montage of overlapping, interfering echoes, impossible to decode. What bats need are pulses of sound, and they must be brief: if two moths are a centimetre apart, then, to produce separate echoes, a sound pulse must have passed beyond the closer one before reaching the second, and hence must be less than a centimetre in length, and therefore less than 30 microseconds in duration.

21.　Ripples vs waves.

It is impossible for bat or machine to produce a short-duration pulse with a single wavelength, unless that wavelength is much shorter than the pulse itself. Instead, bats produce very sharp clicks. Thanks to Fourier analysis, we know that such a click is equivalent to a mixture of waves of diflerent frequencies (Figure 12), and the sharper the click (that is, the quicker it rises from silence to maximum), the wider the range of frequencies produced—and thus the more accurately distances can be judged.

So is the ideal sound for a bat an extremely short-duration click? Not really: a short click means a low-energy one, which cannot travel far without fading. Humans faced this problem themselves in designing radar and sonar systems, and cracked it in 1960 with the invention of frequency sweep. This utilizes a relatively long pulse of gradually increasing frequency. The fairly long duration allows plenty of energy and hence range, and the changing frequency means that echoes from objects at diflerent distances are distinguishable by the diflerences in their frequencies.

When bats use this frequency sweep technique, the pulses last about 2 to 3 milliseconds, during which the frequency falls (rather than rising as in the human version) by about an octave. The more such pulses a bat makes, the more information it receives, and so it modifies its click rate depending on the challenges it faces, from around ten pulses a second when cruising to as high as one hundred in a half-second period, when the surroundings become complicated or when nearing prey.

As discussed in Chapter 2, if the object from which a sound echoes is in motion (relative to the sound source), the frequency of that sound will be shifted by the Doppler eflect. Both human and bat sonar systems exploit this to determine the velocities of such objects, with the distinction that, while our systems measure the change in frequency, bats modify the frequencies of their output pulses until the echoes they hear are the same as they would be, were the target stationary (so, for example, if a bat is approaching an object, the frequency of the echoes from that object rise, so the bat reduces the frequency of its output by an amount that brings the frequency of the echoes back down to match the bat’s original frequency).

A different kind of sound

Box 9

Diffraction of sound through an opening: (plane) angle of spread, θ, is given by sin(θ/2) =λ/d; diameter of the opening d, wavelengthλ.

Another physical diflerence between ultrasound and the audible range is that ultrasound readily forms beams, which is of considerable advantage to bats. An 80 kHz tone passing through a 1 cm opening will form a conical beam which spreads to about 90 cm wide at a 1-metre distance (Box 9). In bat species which project their ultrasound through their nostrils, interference eflects between the two sources mean that the beams are narrowed further. Not only does this concentrate the acoustic energy, so allowing greater distances to be probed, it also reduces the number of distracting side-echoes. A 2015 study of bats approaching drinking pools showed that the bats opened their mouths wider when they were close to the water, presumably for this latter reason.

Evolution could no doubt have furnished bats with generators and detectors that work at even higher frequencies, but the absorption of such sounds by the air is an insuperable barrier. Experimentally it has been determined that, at the sort of conditions bats love best (25°C, 50 per cent relative humidity), 100 kHz sounds are absorbed by 3 dB per metre: that is, they fall to about half of their original intensity over that distance. Conversely, 30 kHz sounds are absorbed only at the modest rate of 0.7 dB per metre, which means their intensity falls by about15 per cent. (The increase of sound absorption with frequency is the main reason why thunder sounds less crackly and more booming as distance increases.)

Box 10

Why does this happen? Air is composed of molecules, all randomly drifting around at a range of velocities and frequently bouncing ofl each other. On a hot day, these velocities increase—in fact, temperature is just a measure of the velocity of a large group of molecules (Box 10). A sound wave is a sequence of alternating high and low pressures moving through the air, so at any given location reached by the sound, the air molecules will first bunch up close together and then spread apart, before bunching together again. When molecules bunch, they slow down, just as a people hurrying through a crowded station concourse in diflerent directions go slower the more of them there are. When this happens, the energy of the molecules changes form: although they move around more slowly (their kinetic energy falls, in other words), they spin and stretch more (so, their internal energy increases).

An analogy might be to pairs of cricket balls linked by strong steel springs—a fairly accurate model of a diatomic (2-atom) molecule, like the molecules of nitrogen (N2) and oxygen (O2) which together make up 99 per cent of our atmosphere. As soon as the compressional part of the sound wave has passed and the molecules move apart again, they spin and stretch less, and move around faster once more. Play middle C, and the sound waves produced force the energy to switch back and forth between the kinetic and internal forms 262 times a second.

But, if the frequency is sufflciently high, the time available becomes so short that a molecule cannot complete the conversion of internal energy to kinetic energy quickly enough before it is time to reverse the process. As a result, the velocity of sound falls and the sound wave dies away rapidly. The actual frequency of the sound wave at which these changes begin depends on the medium, being far higher in solids and liquids than in gases. Other properties of the medium, especially viscosity, also contribute to this eflect.

This phenomenon is very useful to us: as the ultrasound wave fades, its energy spreads through the medium, heating it up. This heating eflect has many applications, including the internal warming of body tissues to improve blood ffow or to treat damaged muscles and joints.

Medical ultrasound

Ultrasound has many other medical uses too, both in the areas of scanning (almost everyone in the developed world is now scanned before birth) and treatment (like the removal of dental tartar, where 25 kHz sound is used in conjunction with a water jet). Unlike many other medical treatments, ultrasound can be switched on and ofl instantly, often requires low-cost technology, and usually involves minimal patient preparation. The fact that ultrasound generators are relatively portable and need little ancillary equipment means that they can be used outside the hospital environment, from diathermy (deep-tissue heating) units found in many gyms to wound cauterization systems, deployed on battlefields to save the lives of soldiers who would otherwise perish from blood loss.

Ultrasound has been used for the treatment of tumours of many kinds—including otherwise inoperable brain cancers—by a technique called either HIFU (high-intensity focussed ultrasound), or HITU (high-intensity therapeutic ultrasound). As well as destroying cancerous tissues by heating them up (to about 90°C over regions about the size of a grain of rice), ultrasonically induced bubble formation in tumours has been used to make them more susceptible to chemotherapy. The main challenge in the use of ultrasound for treatments like these, which need precise and accurate targeting, is that the paths of the ultrasound beams depend on the densities and elasticities of the intervening body tissues. So, models of body parts made from artificial tissue-mimicking materials are used to calibrate and programme the equipment.

A more straightforward use of ultrasound in medicine is lithotripsy, in which high-power pulses simply pound kidney stones in situ, smashing them to pieces small enough to pass out with urine.

Scanning with ultrasound

One of the best-known uses of ultrasound is the scanning of foetuses. Since the diaphragm of a conventional loudspeaker could not move quickly enough to produce the megahertz frequencies required, a piezoelectric projector is used instead. Gel is applied to the abdomen, so that there is no air layer to reffect or absorb the sound. The ultrasound waves bounce ofl interfaces between media with diflerent impedances, such as bone/muscle, or skin/amniotic ffuid. Hence, by measuring precisely how long it takes for echoes to return from these interfaces (taking account of the diflering sound velocities in each medium), their positions can be worked out. By moving the beam around, a detailed three-dimensional map can be calculated, and converted to a real-time video image.

Like most loudspeakers, a piezoelectric transducer is reciprocal—supplied with an oscillating current it produces a sound wave, and, conversely, when struck by a sound wave it generates an electrical signal. So, in a foetal scanner the scanner head acts as both the source of the ultrasounds and the detector of their echoes.

With very high-frequency sound beams relatively sharp images can be made: a 1 MHz (one million hertz) signal can image details at the millimetre scale, the exact value depending on the acoustic properties of the tissue bring viewed, and many scanners now go up to 15 MHz, or 50 MHz for eye and skin scanners.

But even this is paltry compared to the eight billion hertz signals generated by an acoustic microscope, which can consequently resolve details as small as 0.03 micron. Unfortunately, such high-frequency sounds are absorbed by almost all media before they have travelled a single millimetre, and the only useable exception is liquid helium—which will boil if not kept colder than 5 kelvin (fi268°C). The elaborate cooling systems required mean that acoustic microscopes are not cheap: the reasons they are used at all are that they can probe under the surfaces of samples, and that some materials which are hard to distinguish from their surroundings visually are highly reffective to sound.

Outside the medical field, one of the commonest diagnostic applications of ultrasound is in the detection of ffaws and cracks—in railway lines, for instance. To locate these, a series of sound bursts is sent along the object to be tested. In pulse-echo mode, the transmitter and receiver are positioned together: if a ffaw is present, the pulses are reffected and their arrival times indicate the ffaw position. In transmission mode, the detector and transmitter are separated and any changes in the pulses in transit indicate the presence of inhomogeneities in the test object. This approach can also be used for the determination of mechanical stresses in solids: because the elastic moduli of materials alter when stressed, their sound velocities also change locally.

The power of ultrasonics

Though thermal eflects of high-power ultrasound have many applications, it can also deliver its energy through mechanical eflects on the medium. The snapping shrimp provides a rare natural example of the use of the shocks and stresses made by ultrasound (together with audio frequencies) to kill. The loud and sudden click produced when the creature snaps its claw includes frequencies up to 200 kHz, powerful enough to kill or stun both prey and would-be predators alike.

Whether utilized by shrimps or humans, the mechanical power of ultrasound is usually delivered through cavitation, which is the formation and violent collapse of tiny bubbles. Almost all liquids contain such bubbles, made either of their own vapour or of air. When the pressure of a liquid falls, these bubbles grow (which is the cause of the frothing when a pressurized container of fizzy drink is opened). Since a sound wave is a sequence of low and high pressures,

it causes bubbles to swell and shrink rapidly, and at high powers and frequencies the bubbles pulsate so violently they disrupt and implode, suddenly releasing their vibrational energy in the form of heat. The temperatures involved may exceed those on the surface of the sun, and can make the liquid glow with light (a phenomenon called sonoluminescence, which snapping shrimp also initiate).

Because the energy appears only within tiny volumes, the body of liquid does not get particularly hot. But the bursts of highly concentrated energy can be used to initiate chemical changes (sonochemistry), and to clean and sterilize submerged objects such as medical instruments. Cavitation-based cleaning is most eflective in the 20 to 50 kHz range. At higher frequencies, ultrasound also causes agitation of liquids, which dominates cleaning eflects in the range 100 kHz to 1 MHz. In practice, ultrasonic cleaning baths use both eflects.

High-power ultrasound (without cavitation) is routinely used for ffuxless soldering of printed circuit boards, in which the electrically heated tip of a soldering iron is vibrated at ultrasonic frequencies. At still higher powers, fine wires can be welded together, heated from within by the friction caused by ultrasonic vibrations. A great advantage of this is that, since the heating eflect is restricted to the insonified (sound-filled) wires, the ultrasound does not heat up nearby components. Many other materials are welded ultrasonically; the frequency chosen depends on the size of the parts to be welded, from 60 kHz for the smallest elements to 10 kHz for the largest.

High-power ultrasound can even be used to levitate small objects. Although the forces involved are feeble, in microgravity environments such as the interiors of space stations they could be used to hold delicate instruments in position during assembly, or to prevent highly reactive chemicals from coming into contact with anything.

The outer limits: phonons

At extremely high frequencies, sounds behave in ways radically different to those with which we are familiar: just like electromagnetic radiation, in which the highest frequencies behave much more like particles than waves (hence the individual clicks of a Geiger counter in registering the presence of gamma rays), the highest-frequency sounds behave like particles called phonons. Their existence was discovered indirectly: by the late 19th century, it was known that a specific amount of heat was needed to raise the temperature of a substance by one degree. The specific heat of water is higher than that of, say, oil, which means it takes longer to boil a kettle of water than a kettle of oil.

Gases and solids have specific heats too, but there was a puzzling anomaly: increasing the temperature of a solid takes about twice as much heat energy as raising the temperature of the same amount of the same substance in its gaseous form. This means that solids (and certain liquids) must have a means of storing heat which is unavailable to gases. The mechanism is vibration: a molecule in a solid can oscillate around its equilibrium position, like a pendulum. But, unlike a pendulum, an oscillating molecule cannot slow gradually. The laws of quantum mechanics mean that it must jump from a rapid to a slow oscillation, and when it does so it transfers a phonon of vibrational energy to another molecule. The ways in which solids conduct heat and electricity can be accounted for by the behaviour of phonons.

Infrasound

As the lower frequency limit of human hearing is approached, the intensity of just-audible sounds increases: a pure tone that is just audible at 20 Hz is nearly 300 million times more powerful (all other things being equal) than a just-audible tone at 4 kHz. Such powerful low-frequency sounds are rarely encountered in air, but solid-borne versions are commonplace on construction sites, in underground stations, near motorways, and in seismically active areas, where they can easily be felt.

Low-powered airborne infrasounds, on the other hand, surround us all the time. They are even generated when we walk: the cyclic air pressure changes at our ears due to our up and down head movements constitute an infrasound wave at around1 Hz. Sea waves generate infrasounds at around 0.2 Hz. The lowest-frequency natural sounds of all come from high in the air and deep underground: both aurorae and volcanoes produce infrasounds at around 0.01 Hz.

One of the key characteristics of infrasound is that it can travel far further than audible sound, through sea, ground, or air; in air, infrasounds from thousands of kilometres away (made, for example, by erupting volcanoes) can easily be detected, though not usually by microphones. Just as we feel, rather than hear, infrasound, so we use specialized barometers to measure it. Infrasound can also be detected by means of the temperature changes it causes.

There is evidence that infrasound has a range of eflects on humans not shared by other types of sound, including enhancement of emotional responses: when a concert of modern classical music was accompanied by infrasound, there were far more people who hated the music, and far more who loved it, than at a concert of the same music with no infrasonic accompaniment. Other experiments have shown that drivers exposed to infrasound can get very tired very quickly, and it has even been suggested as a cause of supposed hauntings, partly through direct emotional effects and partly through vibration of the eyeball, giving rise to visual disturbances.

Infrasound has proved to be a reliable means of detecting bolides, which are meteoroids that explode in ffight. Since they travel supersonically through the atmosphere, they generate infrasound-rich sonic booms as they fall. More infrasound is generated when the bolides explode. All these airborne infrasounds generate solid-borne versions when they reach the Earth, and surface waves are made when the bolide fragments crash to Earth. The frequencies, timing, and amplitudes of all these sounds can be combined to give a detailed analysis of the path, motion, and energy of the bolide.

However, infrasound is usually of more consequence to us when it travels through the Earth—and ultrasound really comes into its own underwater. It is to these media that we turn in Chapter 7.