WO2002016924A1 - Acoustic interferometry method and device - Google Patents

Abstract

A standing wave interferometry analysis method and sensor (20) for characterizing physical properties of fluids, suspensions and emulsions at ultrasonic frequencies is disclosed. Standing wave features, peaks and valleys are created by continuously changing the ultrasonic frequency by small intervals, between the two ultrasonic transducers (22, 24), which define a sensing zone (30). The transducers (22, 24) are located a known distance apart, for operating in their near field region. Standing wave features are analyzed for frequency location, amplitude and frequency width. The temperature is recorded at each frequency interval. Data from one or more standing wave features are used for calculations of viscosity, density, particle concentration and sound velocity. More accurate particle concentration data is obtained by repetitively scanning the same standing wave peak and measuring resultant frequency shift, caused by particle concentration at the receiving transducer (24). The method is applicable to sub-micrometer particles and measuring the carbon concentration in used engine oil.

Description

ACOUSTIC INTERFEROME RY METHOD AND DEVICE

CROSS REFERENCES TO RELATED APPLICATIONS: This application claims the benefit of U.S. Provisional Application No. 60/227,081, filed August 22, 2000.

BACKGROUND OF THE INVENTION

FIELD OF THE INVENTION

This invention concerns an acoustic interferometry method and device for measuring the solids content, fluid viscosity, and other parameters derived from sound velocity changes in fluids_/ emulsions and suspensions. It may be implemented in online/ in-line, laboratory and Portable applications.

DESCRIPTION OF THE PRIOR ART

The need for monitoring properties of fluids, fluid mixtures, emulsions and suspensions has been apparent in many different industries and applications for many years. This Prior Art description is divided into sections according to the parameters that can be measured with this invention. SOLIDS CONTENT AND PARTICLE SIZE

The measurement of the solids content of a suspension can be achieved in many ways. One simple method is to filter out the solid material from a known amount of suspension and then weigh the amount retained by the filter used. The accuracy of this simple gravimetric method is dependent upon the efficiency of the filter used for removing the particles. If the solid particles are smaller than the pore size of the filter, the particles will not be retained and inaccurate results will be obtained since some of the particles pass through the filter.

Another method used for the determination of solids content involves the measurement of the turbidity or haziness of the suspension. A light obscuration or a light scattering method can be used for this determination is particularly applicable to small particles and low solids content, since large particles may sediment through the measurement beam; and high solids concentrations can obscure the light beam completely, or produce excessive scattering at chosen scattering angles. The detection of backscattered light can be used for more concentrated sample systems, but care must be taken with the cleanliness of the sample cell / sensor interface. One difficult application involves the measurement of dark- colored or opaque systems. It is a particularly difficult application, if the solid / fluid system comprises sub- micrometer sized particles dispersed in an opaque liquid. An example of such a system is the measurement of carbon (fuel soot) in engine oil. In this application, the oil is usually viscous, opaque and also contaminated with materials other than the carbon itself. This application has been performed for many years, but typically laboratory analysis is performed on a sample taken and bottled directly from the engine. The methodologies used for these laboratory analyses are the accurate, Thermal Gravimetric Method, or more usually Fourier Transform Infrared spectroscopy.

An implementation of infrared signal interpretation was disclosed in US patent 4,570,069, 1986 by Gager. This patent details an on-board oil quality monitor that is based upon the transmitted infrared radiation through the oil sample. The oil- quality is determined from the transmitted radiation. An oil change is recommended when the signal transmission falls below a preset threshold limit. The transmission of infrared radiation is known to depend upon many factors and the presence of many different chemicals. Infrared radiation is unsuited for the accurate determination of solids alone, unless referenced against a known uncontaminated reference oil and preferably under laboratory conditions.

In US patent 4,478,072, 1984, Brown details an apparatus for determining the concentration of solids dispersed in a liquid. An example details the concentration measurement of coal in oil. Another use for this Apparatus would be to determine the concentration of carbon in engine oil. This apparatus comprises an ultrasonic pulser that sends ultrasonic pulses into the sample and reference material, such that the reflection coefficient of the interface between them is recorded. The reflected signal depends upon the difference between the reflection coefficient of reference material and sample. Several disadvantages to automotive applications are apparent for this apparatus. The results are dependent upon both the sample and reference material being at the same temperature. Differences in thermal conductivity between the sample and reference material make temperature stabilization in an on-line or continuous use operation difficult. In addition, the bulk property impedance as measured is dependent upon many parameters including temperature, density, surface contamination, and solids content. The reference material as detailed is not suitable for use at the elevated temperatures found in an internal combustion engine. US Patent 3,779,070, 1973 by Cushman et al discloses an acoustic method for the determination of both solids content and particle size. Two ultrasonic signals at two different frequencies are passed through a flowing stream of slurry and the received acoustic signal attenuation interpreted to yield the solids volume content and the geometric means particle diameter. This method and apparatus was developed for relatively large particles in water, the examples showing particle size distributions substantially larger than 1 μm in diameter.

US patent 4,412,451, 1983 by Uusitalo et al details an ultrasonic based method and apparatus for the determination of the average particle size of slurry. This patent discloses the measurement of aqueous systems; and it measures the scattering attenuation of one or two transmitted ultrasound beams at two different frequencies directed within the slurry. The transmission of one beam and the scattered intensity of the other are recorded. The method eliminates the attenuation changes due to the density of the water in which the particles are dispersed. The method shows how ultrasonic signal attenuation varies with particle size and frequency within the range of 1 to >2.4 MHz. A further implementation of an ultrasonic oil condition detecting apparatus was disclosed in US patent 4,785,287 1988 by Honma et al in which an ultrasonic pulse is reflected from a reflector plate and changes in the waveform involving time are translated into particular condition of the oil. The inventors did not disclose the measurement of sample temperature. Very large background attenuation is caused by changes in the sample temperature.

A further method, disclosed by Yasuhara et al in US Patent 4,646,070, 1987, determines the deterioration of engine oil by measuring its dielectric constant. The disadvantage of this and other methods that measure the dielectric constant of oil is that the readings are generally responsive to chemical changes as well as contamination by solids and liquids. Separating and analyzing the gross property dielectric data is considered a difficult task. US Patent 5,262,732 1993, by Dickert et al addresses this further by obtaining metal wear particle data by extracting information while measuring the dielectric constant of oil under the influence of a magnetic field. US patent 5,540,086 1996 Park et al also details an oil deterioration sensor based upon the measurement of the dielectric constant of the oil. US patent 5,604,441 also measures the dielectric constant of the oil and is able to discriminate between wear and other particles much greater than 1 μm in size by a magnetic separation system.

Another ultrasonic method for the determination of both solids concentration and particle size distribution is disclosed in US Patent 4,706,509, 1987 by Riebel. This method and apparatus measures the absorption of ultrasonic waves at multiple frequencies; a different frequency for every size increment in the particle size distribution required is also disclosed. Riebel states that ten frequency steps are adequate to cover the entire particle size spectrum. The frequency steps are chosen such that the wavelength of the highest frequency corresponds to or is smaller than the diameter of the smallest particle to be measured; and the lowest frequency corresponds to or is greater than the diameter of the largest particle to be measured. This method uses a reference cell containing a reference fluid free of particles. Data is obtained by analyzing the fluid over a wide frequency range. The solids concentration is determined from the sum of all frequency data. This method does not address specifically the problems associated with on-line use. Although a reference sample is used, no consideration is given to the variation in signal attenuation caused by temperature variations, both between the reference and sample, and within each analysis period. In on- line applications and in on-engine use, it is impractical to utilize a reference fluid, especially if a dynamic system and other components or contamination are present.

Further improvements in ultrasonic technology lead to the more accurate determination of particle size distribution and concentration in suspension detailed in US patent 5,121,629, 1992 by Alba. This laboratory bench instrument determines the multi-frequency attenuation spectrum of the dispersion. Subsequent mathematical matrix inversion treatment extracts the particle size distribution data within a size range dependent upon the instrument set-up parameters. This method is able to discriminate between size data for sub-micrometer particles. One important factor with this system is that to measure the small attenuation changes that typically occur (due to the presence of particles) the sound must travel distances much greater than the wavelength of the sound to provide sufficient measurement sensitivity. US patent 5,569,844 by Sowerby in 1996 discloses the use of additional ultrasonic transducers and gamma-ray transmission gauges. This laboratory system appears unsuited to the on-line analysis or on-engine measurement of soot in engine oil. A more recent approach for determination of soot in engine oil was disclosed in US patent 5,914,890 in 1999. Sarangapani et al details a mathematical modeling method based upon engine operational parameters. This method approximates the solids content of the oil and is solely dependent upon the reliability of the modeling used.

VISCOSITY AND DENSITY

US Patent 4,117,719, 1978 by Simon details a Densitometer Apparatus and Sonic Veloci eter that measures the resonant frequency of a piezoelectric transformer interfaced with a membrane to seal against the fluid under test.

The measurement of viscosity from ultrasonic signal attenuation was disclosed in US patent 4,559,810, 1985 by Hinrichs et al . The inventors determine the dynamic viscosity of the specimen of polymeric resin subjected to time varying temperature changes. Ultrasonic Continuous-wave ultrasonic frequency pulses of known amplitude are transmitted a distance through the sample and the resultant attenuation determined.

In US patent 4,721,874, 1988, Emmert measures viscosity with an ultrasonic transducer operating in shear mode coupled to the test fluid with a polymeric layer slightly thicker than one- quarter wavelength. US Patents 4,862,384, 5,365,778, 5,698,773 and 6,227,040 Bl also describe ultrasonic methods for determination of dynamic viscosity; and in US Patent 5,798,452, 1998 Martin et al discloses a textured-surface quartz resonator that determines the dynamic viscosity and density of fluids. The above disclosures of prior art use low cost ultrasonic oscillators for the determination of fluid viscosity.

EMULSIONS

US patent 5,599,460, 1997, discloses a Water / Glycol Sensor that measures the dielectric constant of the oil sample in a water / glycol concentrated zone.

INTERFEROMETRY

There are many published documents regarding interferometers and it is well known in the art how sensitive such devices are to detecting changes in the signal or the parameter under test.

Acoustic interferometry methods that rely upon the change of the length of the sound path to locate wave interference or standing-wave location have been used for many years and are described in college text books with respect to speed of sound in air. Apparatus or devices that utilize acoustic interferometry techniques have been use to determine fluid parameters such as density. US patent 5,359,541, 1994 discloses an ultrasonic resonance interferometer that simultaneously subjects the test sample and a reference sample, both placed in resonators, to a band of ultrasonic frequencies. The data is extracted from the Fast Fourier Transform of the signal that contains many resonant frequencies, and is used to determine specific gravity and solute concentration.

US Patent 5,533,402 details a method and apparatus for measuring acoustic parameters in liquids using cylindrical ultrasonic standing waves. This device measures the sound velocity and attenuation from the oscillation of a cylinder. A reference cell is used to provide relevant data for extracting fluid parameters .

Belonenko et al in US patent 5,804,698, 1998 disclose a similar method and system for measuring fluid parameters by ultrasonic resonance. This system also uses two or more ultrasonic resonators, one of which is filled with a reference fluid and determines the resonant frequency by varying the phase of a selected frequency supplied to the transmitting ultrasonic electro-acoustic oscillator. The fluid temperature of test samples can vary with the location of its measurement and this is especially true in an Internal Combustion Engine (ICE), where the temperature can fluctuate rapidly, according to engine load and ambient conditions. In the case of thermally insulating fluids, such as engine lubricating oil, temperature gradients are a particular problem to many test methods. The sample temperature vastly influences acoustic analytical methods, and even small changes or gradients in temperature cause significant effects. Changes in the sample temperature often cause larger changes in the sought after data than the changes obtained by the parameter being measured. In the laboratory, it is straightforward to control the temperature of the sample under test with a water bath or similar means; however, outside of the laboratory the situation is more difficult. The more thermally insulating the fluid, the more that temperature gradients and fluctuations are likely to be present and this is particularly true in lubricating oil systems found on internal combustion engines .

Sound propagated in fluid can move as traveling longitudinal acoustic waves. These may be considered as pressure pulses that are parallel to the direction of propagation and that move longitudinally through the system with a wavelength, λ, at

velocity, v and of frequency, f.

The velocity of sound in a pure liquid is given by the Woods equation (1), where B is the coefficient of compressional

viscosity of the liquid and p, its density.

The Woods equation implies that sound velocity is independent of frequency and power level and depends only on the elasticity (coefficient of compressional viscosity) and density of the material. Sound propagation in fluid mixtures, colloids, and suspensions is however more complicated. The sound velocity is modified by the presence of particles. The variation of sound intensity caused by absorption or attenuation is both very temperature and frequency dependent.

It is assumed that the sound velocity is independent of the sound power levels if low sound power levels are used and that sound velocity is also independent of frequency.

Sound can be considered a series of pressure pulses that as they travel create an in-phase temperature variation. For low power sound waves, the pressure pulses generate very small temperature fluctuations of the order of <1 mK.

For mineral oil, the velocity is approximately 1625 m/s

at 20°C (from tables.) The velocity of sound varies with temperature such that a negative linear coefficient describes

the change. A typical coefficient value, δv/δt is -2.8°C (for olive oil from reference tables in Kaye & Laby) .

Apparatus for measuring acoustic properties generally comprises a piston like source that issues uniform symmetrical waveform pulses; piezoelectric ceramic transducers for example PZT fall in to this category. Pulses from this type of sound source are transmitted through liquids in plane waves if operating in longitudinal mode. The area close to the sound source is called the near field and is described by a

cylindrical area r²/λ long and 2r in diameter.

Povey (p 20-22) discusses that: ''although the wave- front is nearly plane within the near field the amplitude varies wildly with distance and so the near field is generally unsuitable for amplitude measurements because the signal amplitude is so sensitive to the distance from the transducer/' Beyond the near field lies the far field in which the wave-front is no longer plane and the signal amplitude is an exponentially decaying function of distance. In the far region, the wave- front spreads out tending towards a spherical shape and its diffraction contributes approximately 1 dB of attenuation per

distance of r²/λ. Consequently, to achieve good performance from a measurement system, prior art teaches that the ultrasonic wavelength should be made much smaller than the sample containing dimensions and much greater than the size of any particles dispersed within the sample (Povey pl04) .

The attenuation of a sound signal traveling through a medium is proportional to its frequency and so it is normal practice to use higher rather than lower frequencies to increase the attenuation caused by the signal of interest.

Various methods have been used for the measurement of sound velocity and absorption; some discussed herein as prior art. Most those of those have been adopted to endeavor to minimize the reflection of sound waves and creation of standing waves that might interfere with the measurement. Many methods use ultrasonic transducers that work at frequencies within the range of 3 MHz to >100 MHz. This range is chosen to both minimize standing wave interference (the bottom limit) and enhance attenuation of the desired signal. For an ultrasonic system within the range given above, this leads to a transmitter / receiver separation of 10 mm to approximately 100mm and larger depending upon frequency and requirements .

The thickness and diameter of piezoelectric transducers determine their resonant frequency and mode of operation. Their thickness is chosen according to the frequency range of operation required. Their diameter (and manufacturing parameters known to the transducer manufacturer) determines whether the transducers will oscillate in the longitudinal mode or shear mode.

The presence of particles within a suspension or emulsion leads to a change in the sound velocity. Particles also cause attenuation of a sound wave as it passes through a suspension or emulsion. Prior art references herein detail equations relating attenuation to particle size and solids content.

Propagation of sound through an inhomogeneous material such as used engine oil involves scattering of the sound. Particles within the liquid sample cause the scattering of sound. The quantity and size of these particles influences this scattering. Multiple scattering processes contribute to the net change in the above parameters seen when the solids content of the sample changes. Forward scattering results in a phase shift causing a shift in the sound velocity. This change in sound velocity can be described by equation 2:

= - (l + aΦ + βΦ² + χΦ³) (2) v v_x

where v is the velocity in the dispersion; i is the velocity in the fluid; Φ is the volume fraction of the particles.

The coefficients α, β can be determined from known dispersion concentrations and the assumption of similar particle size distributions. The coefficient χ is small and so is ignored. This equation can be used for the determination of the concentration of particles present in well-defined systems, such as carbon in engine oil. In practice described herein, a calibration that correlates particle concentration with the observed frequency shift (change in sound velocity) is performed. Acoustic standing waves are created between two walls (or transducers) when a sinusoidal wave is reflected from a wall. The following equation (3) , given in many college textbooks, describes a general equation for describing any standing wave n, over a finite distance:

Ψ_n (x, t) = A_nu_n (x) cos { ω_nt + φ_n) (3)

where Ψ is the wave displacement, t is the periodic time, at + φ is the phase angle and φ is the phase constant. The space

functions u_n (x) are the Eigenfunctions of the system.

The standing wave described above is stationary, but changing the phase would move its physical location between the two walls or transducers. The standing wave repeats itself in a distance λ, the wavelength. The number of standing waves and their positions are influenced by the frequency and velocity of the wave. At positions corresponding to kx = mπ, where m = 0, +1, ±2...), Ψ = 0. Midway between these positions (nodes) are the antinodes where the amplitude is greatest. SUMMARY OF THE INVENTION

An acoustic interferometry method and device for measuring solids content, viscosity and the density/ elasticity of fluids, suspensions and emulsions. Standing waves, created by continuously transmitting sound pulses at discrete frequency steps through the test sample located between two piezoelectric transducers inside a chamber or tube, are analyzed for peak or valley amplitude, frequency location and width. One or more features are located from a ^λλ spectrum" of available standing waves, corresponding to wavelength fractions or multiples varying according to: the frequency chosen, the transducer separation distance and the fluid parameters. The standing wave features are measured by the receiver during the continuous stepwise frequency scan. Each step size is chosen to facilitate accurate location of the chosen standing wave peak(s) or valley (s) . A temperature sensor is used to provide continuous temperature data throughout the measurement period. A complete scan of one or more peaks or valleys is performed at least once to provide viscosity and density-elasticity data. The viscosity can be determined directly from the signal voltage amplitude and the density-elasticity data from a peak or valley frequency location. The peaks can be considered to represent the maximum transfer of energy to the receiver transducer; whereas, the valleys represent the minimum transfer of energy. Sound velocity data is best obtained from the location of a valley at a frequency corresponding to one wavelength between the transducers since minimum energy transfer occurs there, thus minimizing temperature gradients. Velocity data also can be obtained from differences in peak or valley locations. The concentration of insoluble liquid components, in the form of emulsions also can be determined especially in a system not contaminated with solids, for example, water in diesel fuel and hydraulic fluid. In addition, solute concentration, or liquid mixture concentrations can be determined by analyzing standing wave feature frequency locations .

The spectrum of standing waves obtained from this method comprises peaks and valleys of different voltage amplitude that coincide with fractions of a wavelength, odd and even numbers of wavelengths and those created by influence of the reflections. Relative density data can be obtained from a peak or valley location and is ideal for monitoring concentration changes in mixtures of liquids or solute concentrations. Better density data can be obtained from multiple peak and valley location information, since for some systems, liquid elasticity can be eliminated. The solids content can be obtained by repeating the frequency scan at least one more time. To measure solids, a narrow frequency range is chosen so that only one peak is scanned. During each scan, heat is transferred to the sample from the acoustic signal, so that the temperature is recorded continuously. During each scan of a peak, solid particles are transported toward the receiver. This particle concentration movement is completed after two or more repetitive frequency scans of the same standing wave peak. The increasing particle concentration at the receiver surface causes a frequency shift, or apparent phase velocity change. This frequency shift is proportional to the solids content. The rate of change of frequency shift is dependent upon: the sample temperature, the particle density difference between the particles and fluid, the particle size, the fluid viscosity and system parameters, such as the applied power, transducer separation distance and the sample confinement area. An equilibrated state is obtained after several scans, following which no additional frequency shift is noticed, except for that caused by temperature changes alone.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. la shows a perspective view of the ultrasonic sensor head;

FIG. lb shows a cross-sectional view of one embodiment of the ultrasonic sensor head;

FIG. 2 is a cross-sectional view of a second embodiment of the sensor head;

FIG. 3 shows a general schematic of the sensor device;

FIG. 4 is a graph of the standing wave features; valleys and peaks (nodes and anti-nodes) that are generated during a step- wise frequency scan of 15W-40 multi-grade heavy-duty engine oil at a constant temperature;

FIG. 5 is a frequency scan for water measured at one temperature and showing the standing wave features obtained;

FIG. 6 shows standing wave valley data obtained at a temperature

of approximately 84°C for three carbon contaminated engine oil samples; FIG. 7 shows repeat high-resolution scans of one peak generated for engine oil samples containing 0.6% and 4.1% soot;

FIG. 8 shows standing-wave peaks for various mineral oil viscosity standards;

FIG. 9 gives a calibration graph of signal voltage versus Dynamic Viscosity;

FIG. 10 is a plot of Signal Voltage versus Dynamic Viscosity for Standard S60 obtained between the temperature range of 25 and

40°C;

FIG. 11 is a graph of double standing wave peaks used to determine the densities of mineral oil standards; and

FIG. 12 is a plot of frequency versus signal voltage obtained for water in soybean oil emulsions.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

The acoustic sensor device of this invention can be used with fluids, suspensions and emulsions. The invention method can be used with aqueous systems, solvent systems and oils, such as mineral and synthetic engine and hydraulic oils. The sensor device can be implemented in laboratory instrumentation, in online equipment, and in-line vehicle and chemical plant applications, where the temperature of the sample fluid can vary

within the range of sub-zero to a maximum of 125°C.

The preferred implementation of the acoustic sensor 20 of this invention is shown in Figs, la and lb fitted with a rapid response temperature sensor 21. The response rate of a temperature sensor is reported as its time-constant, which is defined as the time taken to reach 98% of reading, following a step change in temperature. The response rate required for the preferred sensor device is less than 10 seconds, more preferably less than 1 second and most preferably a less than 100 milliseconds .

The sensor device uses two piezo-electric transducers 22, 24 that are obtainable, custom built, from many manufacturers and suppliers, for example, American Piezo Ceramics of Mackeyville Pennsylvania. Their chemical composition usually is based upon derivations or mixtures containing Lead Zirconium Titanate and is usually known by the acronym PZT. The electrical connections 26, 28 are achieved through silver or other conductive coating on both sides of the transducer. The transducers are polarized during the manufacturing process and the preferred embodiment of this sensor device has the positive electrode connection wrapped-around to the other side so that both electrical connections are physically made on the electrically negative side. One of the transducers, the transmitter is electrically energized with a continuously oscillating voltage signal that preferably is a sine-wave signal, but can be a triangular wave or a square wave. In the preferred embodiment it does not matter which of the ultrasonic transducers 22, 24 is used to transmit the acoustic pulses. The receiver transducer 22 or 24 receives the transmitted signal that can comprise a standing wave node, anti-node or have no significant slope change.

The two transducers 22, 24 are spaced apart and parallel to each other with their positive electrical connections, facing each other within a tolerance of approximately 10 micrometers across the diameter of the transducer. For the preferred embodiment of the sensor device, it is not necessary to very closely control this tolerance if a simple characterizing step is performed for each sensor device in the form of a ''calibration constant". The closer the transducers are to each other, the less obvious is the effect of parallel misalignment.

The transducer holder 32, hollow chamber 36 with threads 42 are all constructed from an insulating plastic material, for

example, polyetherimide "Ultem"™ available from GE Plastics) . The use of an insulating material such as Ultem is considered essential to prevent electrical shorting of the ultrasonic transducers and also to prevent acoustic interference or ringing that might occur from high velocity sound traveling through metallic components.

Contrary to the prior teaching of Povey and others, this invention uses a transducer spacing of typically less than 10 mm, with the intention of creating standing waves. The transducer spacing creates a sensing zone 30, and is determined from the frequency of operation and the sensitivity of the measurement required. The choice of standing waves at or around a frequency equivalent to 4 or 6 wavelengths gives the best results. The ability of the sensor to detect acoustic changes improves with the number of sound wavelengths that are fitted within the sensing zone. ^'The standing wave peaks become more pronounced since their construction involves multiple reflections .

The transducer spacing is determined from knowledge of the approximate speed of sound in the fluid, the required frequency of operation and the number of required wavelengths for standing wave generation. The preferred embodiment, suitable for the measurement of oils, can utilize a transducer spacing of 1 to 15 mm and preferably a spacing of 2-10 mm and most preferably 4-6 mm.

The transducer spacing distance L can be calculated from the sound velocity in the fluid of interest v, the number of required wavelengths N within distance L, and the frequency of operation f . This is shown in equation (4) :

L = N . v /f (4)

The ultrasonic transducers 22, 24 can be spaced apart a distance equivalent to between 0.5 to 10 wavelengths, but preferably are within a range of 0.5 to 6 wavelengths apart, corresponding to each transducer spacing within its near-field region. The frequency of operation can be 0.1 to 100 Mhz, preferably 0.3 to 10 MHz, or most preferably 0.6 to 1.6 MHz. The two ultrasonic transducers, 22 and 24 are mounted facing each other as shown in Figs, la and lb. One transducer is used to transmit the sound signal and the other used for receiving the signal. The transducers are located such that they are separated by a distance that corresponds to approximately one one-half, one, two or more approximate wavelengths apart. The speed of sound and the available sound frequency range are used to determine the transducer separation distance required.

Referring to Figs, la and lb, the distance between the transducers 22, 24 forms the sensing zone 30 in which the test sample is located. It is an integral part of the transducer holder 32. Connecting wires (not shown) from the transducer electrodes 26, 28 are positioned such that they pass through the center 34 of the hollow housing 36 and out through an exit hole 38. They are sealed or potted with a suitable resin, for example epoxy resin, such that a "fluid tight" seal is obtained. The housing is sealed with the aid of O-ring 40, and screw threads 42. Temperature sensor 21 is located outside of the transducer, periphery, yet still close enough to provide realistic sensing zone 30 temperature readings.

If accurate solids measurement is not required, then the acoustic sensor device 20 shown in Fig. la can be used and located inside of the flowing stream, without the need for a housing 44. However, its location must provide for a homogeneous supply of test fluid and must also avoid the presence of non-intentional contamination, such as air.

For accurate solids determinations, it is necessary to avoid influence from outside of the sensing zone. For this reason, the sensor is located inside of a chamber, housing or tube structure 44, with inlet 46, isolation valve 48, and outlet 50. The housing 44 is designed to minimize the sample test volume located outside of the sensing zone 30. Minimizing this excess sample space helps to reduce the effects of fluid and particle diffusion that can occur during the sample test period. The inlet 46 and outlet 50 should preferably be located to ensure that air is not trapped within the housing 44.

The presence of air bubbles can significantly affect the data. Because of its affect, it is however straightforward for the sensor control software to detect the presence of such temporary contamination and prevent bad data from being reported if found.

Fig. 2 shows an alternative embodiment of the sensor 20' suitable for measurement of solids content, viscosity, density- elasticity data and sound velocity data. Electrical connections 52 (wires not shown) from transducers 22 and 24 pass through channels 54 and pass out through channel 38 together with the electrical connections 56 from temperature sensor 21. The test sample 58 is introduced into the sensing zone 30 through test sample inlet 46, located in sensor body 60. This embodiment requires a small sample volume, sufficient only to fill the sensing zone, typically 0.5 ml.

The preferred embodiment of the method provides for scanning the frequency throughout a range of 0.6 to 1.6 MHz and provides for the frequency to be changed in 100 Hz steps or multiples thereof. An alternative frequency range can be selected by adapting the frequency synthesizer circuitry / phase-lock loop control (PLL) 66. Example ranges are 0.6 to 3.2 Mhz or 1.2 to 6.4 Mhz. Adapting the PLL control to a different frequency range is a simple process to those experienced in the design of frequency control methods. The system can be designed to change frequency ranges automatically on command by the microcomputer 62.

The piezoelectric transducer size is chosen according to the power requirements. The larger the transducer, the higher the power is required to make it oscillate. A power requirement of 3 Watts or less is adequate, but lower power values are used to minimize heat transfer to the test sample. The preferred embodiment uses a transducer diameter of 6 mm and an oscillator power output of approximately 1.5 Watts. The ultrasonic transducer thickness is chosen according to the required operating frequency range. It is preferable to select a transducer thickness that corresponds to a resonant frequency that is above the required frequency range of operation, since a smooth, flat response, devoid of resonant peaks, gives good performance. The ultrasonic transducer output has been shown to be reproducible and consistent, if operated below its natural resonant frequency. Above this natural resonant frequency, multiple resonance peaks can occur; and their location is temperature dependent. For a frequency range of 0.6 to 1.6 MHz, a transducer thickness of 2 mm is suitable.

The method of analysis determines the frequency location of one or more standing waves generated between the transducers . Incrementally increasing, or decreasing the frequency in steps, between two pre-determined values, creates the standing wave or waves required. Frequency steps of 1kHz apart are used to find an approximate peak location for the automatic determination of the correct frequency range for a detailed scan. The size of the frequency steps used to perform sample tests is preferably 300 to 500 Hz apart, or most preferably 100 Hz or 50 Hz apart.

Each standing wave feature, whether peak or valley, is scanned throughout a range of 15 to 40 KHz, which is a range sufficient to fully encompass the location of the standing wave feature (peak or valley) . For some signal interpretation algorithms, described later, it is necessary to scan two or more standing wave features. Multiple peaks can be scanned either by changing the frequency in 100 Hz steps throughout the chosen range, or by changing the frequency in 100 Hz steps, while measuring each feature, and then stepping the frequency in much larger increments, such as 1 KHz or 10 KHz between features. The time between each frequency step is dependent upon analysis requirements, sample and system parameters. The most important system parameter is the phase-lock loop lockup time (the time taken for the frequency to stabilize following a change) . The lockup time is a design parameter known to those experienced in frequency control system design and is typically chosen to be 1 millisecond for frequency steps of 1 KHz. The frequency synthesizer 66 shown in Fig. 3, can control the step size to a resolution of 100 Hz. The frequency synthesizer PLL control 66, uses a lockup time of 10 to 50 milliseconds to stabilize each frequency step of 100 Hz. Other system parameters that determine the time required for each frequency step include, the number of data (signal voltage and temperature) readings taken. In the preferred embodiment of this method, the number of signal and temperature readings taken is chosen according to the mode of operation, the sample reading throughput capacity and the fluid parameter under test. The accuracy and reproducibility of the data improves as the number of sample readings increases. These signal-averaging techniques can help to reduce system noise and are well known to those experienced in data acquisition methods. The modes of operation used in this method are status recording (for updating device status information) , frequency change (to step between features and for automatic peak location) and sample analysis. One hundred individual sample readings are averaged for status recording and frequency change mode operations and one thousand or more are chosen for the sample analysis mode. The time taken to record these sample readings is dependent upon the sampling rate chosen and the throughput of the electronics. A value of 10 KHz is adequate for good performance and is inexpensive to implement. Use of a higher sampling rate will give better performance if cost is not a concern. The total time taken at each frequency step is the frequency stabilization time plus the sampling time; it is approximately 20 milliseconds for status recording and frequency record modes and approximately 110 milliseconds for the sample analysis mode. The total analysis time for analyzing a single standing wave feature, assuming a typical frequency range of 20 KHz, is 22 seconds.

The frequency width of the standing-wave feature is dependent upon system parameters, the sample, and also the temperature change that occurs throughout the sample test scan. The data obtained during a sample test scan is rejected if the temperature variation exceeds predetermined limits. The temperature variation limits are chosen according to the accuracy of the data required. For viscosity and sound velocity related parameters, the intra-run temperature variation should

not exceed 0.1°C, preferably 0.05°C and most preferably <0.01°C. The faster the analysis is performed, the easier is to achieve the stable temperatures required. In practice, it is easy to achieve the stability required. If the temperature variation is too large, the standing wave feature becomes broader in frequency and some frequency shift will occur; the direction will depend upon whether the temperature is rising or falling.

Referring to Fig. 3, a general schematic for the acoustic signal generator and signal interpret means 61 and the sensor device 20, a crystal controlled frequency synthesizer 66, with phase-lock loop (PLL) control and amplifier circuit 69, is controlled by a microcomputer 62, and is used to generate a square wave of precisely known frequency. The microcomputer 62 is provided with a scan data input 64, that comprises user input, temperature feedback 86 from the temperature sensor 21, and signal feedback 88, to determine when (at what temperature) and how (what scan frequency range) to make the measurement (s) .

The choice of electronic components must ensure that the system can set a specific frequency to within 100 Hz with an

accuracy of ±50 Hz and preferably with an accuracy of ±25 Hz. The preferred device of this method utilizes the following main components to form its frequency synthesizer PLL and amplifier system 66: a high accuracy 1.024 MHz 25ppm crystal oscillator, available from Epson Corporation, a Frequency Synthesizer Chip MC14151 (Motorola Corporation) and a Voltage Controlled Oscillator chip 74HC7046, available from several companies, such as the Fairchild Corporation. The square wave signal is passed through an output 68, and is filtered and amplified with a highspeed "rail to rail" amplifier 69 available from many sources. This square wave of known frequency and 50% duty cycle is electronically filtered to remove the unwanted frequency pulses that might be present; these may originate from many sources such as the PLL itself, stray capacitances and power supply. The filtering process is accomplished with this "active filter" known to those experienced in electronics. The filtered square wave is further amplified with an operational amplifier 70, available from many sources, to provide sufficient current gain for driving the transmitter transducer. The voltage from amplifier 70 is increased with the aid of a high frequency transformer 71, (available from many electronic component suppliers) , or by other electronic means, to a value suitable to energize the piezo-electric transmitting transducer 22 via connection wires 72. The applied voltage required most ideally is less than 45 VAC for safety reasons and preferably chosen to be approximately 36 VAC.

The signal, after traveling through the sample 30, is measured by the receiver transducer 24, is triangular or sine wave in shape and passes through connections 74. The received signal is filtered and the DC component removed with a low-pass filter and capacitor combination 76 to produce a clean signal that passes through a variable gain amplifier 78, for example THS7001 available from Texas Instruments of Austin, TX. The variable gain of amplifier 78 ensures that the maximum- signal voltage is within the required specifications. In the preferred embodiment a signal voltage range of 0 to 3.5 volts is used, but this value can be any suitable value for the electronic circuitry used. The attenuated or amplified signal on a line 79 is passed to a peak height determining circuit 80, that comprises a dual operational amplifier voltage /hold circuit that provides the maximum voltage output of the transducer received signal. The maximum voltage output and frequency is recorded by the microcomputer 62 for each step of the frequency scan. Each of the data point sets for each standing wave feature "scan" is passed to a signal interpretation algorithm means 84. The specific signal interpretation algorithm 84 is selected automatically according to the required sample test data. Different data treatments are provided by the algorithm means 84 for each available measurement as detailed herein below, under individual headings.

The transducers 22, 24 used for the preferred embodiment of this invention oscillate primarily in the longitudinal dimension and generate plane acoustic waves that are reflected between the transmitter 22 and receiver 24. The transducers also generate some shear waves, since some of the edges of the transducers are exposed to the sample fluid under test. The proportion of longitudinal to shear waves is not known.

The chemical and electrical compatibility of the transducers with the fluid under test must be considered. Transducers, exposed directly to the fluid, have been shown to be compatible with all serviceable engine and hydraulic oils tested to date. The transducers have operated for an extended time period in mineral oils, synthetic oils, alcohols and aqueous systems. If operation with corrosive or aqueous electrolyte systems is required, the transducers should be protected with a coating, for example epoxy resin, or they can be protected with a metal shield. The method for achieving this protection is well known to transducer manufacturers and those skilled in the art. The characteristics of the transducers will be changed however so that the system characteristics must be known. Such characteristics include the transducer frequency response, the oscillation power required and the transducer spacing required.

This Fig. 3 system, comprising the acoustic sensor device 20, and the acoustic signal generator / signal interpret means 61, measures the standing wave "spectrum" data created by changing the frequency stepwise between two values. It provides phase velocity, sound velocity and attenuation data together with a continuous reading of temperature. The temperature data provides real time information about the transfer of heat during the sample test measuring period and the data provides supporting information about the frequency locations of the standing wave features. The temperature sensor 21 position within the acoustic sensor 20, 20' determines whether the temperature readings it provides are "in-phase" with the detected standing wave. The temperature sensor is one side of the sample test area. If an ultrasonic frequency equivalent to one wavelength is applied between the transducers 22, 24, the

recorded temperature signal appears to be 180° out-of-phase with the receiver 24 signal.

SIGNAL INTERPRETATION ALGORITHMS

The following data is recorded for each standing wave feature of interest: a) The maximum peak height in volts, corrected for the setting of the variable attenuator 78, to ensure that all values are "normalized" to a gain setting of one, or an appropriate microcomputer controlled value. The resultant value corresponds to the boundary value of A (amplitude) in a harmonic wave described by ψ( t) = A cos (ω₀t + φ) ; where ψ is the displacement, t is the periodic time and ω₀t + φ is the phase angle; b) the frequency in Hz at each of the data points; c) The temperature at each of the data points to a resolution

of better than 0.01°C; and

d) the time is also recorded, but this is not a requirement.

A plot of Frequency versus Signal Voltage is shown in Fig. 4. It is a plot of the standing wave valleys (nodes) and peaks (anti-nodes) that are generated during a step-wise frequency scan of 15W-40 multi-grade heavy-duty engine oil. This plot was obtained using frequency intervals of 500 Hz stepping between 300 and 1650 KHz; it took approximately two minutes to measure. This plot is presented for information and should be used for determining the frequency range to encompass suitable features for detailed analysis. The software for the microcomputer 62 is programmed to automatically locate the selected features according to requirements. The software searches for the appropriate maximum or minimum values and the detailed analysis using a step size of 100 Hz is performed over a range centered on the location (s) found. It does not matter whether the scan incorporates increasing or decreasing frequency values.

Fig. 5 is a plot of Frequency versus Signal Voltage for

distilled water, measured at a temperature of 29.3°C. Water has

been well characterized by many workers. Assuming a sound velocity value of 1512 m/s for water at 29°C, the wavelength λ equivalence of the standing wave features for an approximate transducer separation of 4 mm are given in Fig 5: peak 90 = 2.5λ, valley 92 = 2.75λ, peak 94 = 3.0 λ, peak 96 = 3.5λ and peak 98 = 4.0 λ. The measurement of water or other known fluid can be used to accurately calibrate the transducer separation distance 30 and to calibrate its characteristics, since the peak height can be used to calibrate the response of the receiving transducer 24.

The recorded data a, b, c and d is analyzed using the Signal Interpret Algorithm means 84, which determines the required results: solids content, viscosity, density, sound velocity or other required parameter of the test fluid. The data obtained for one or more standing wave features is subjected to further data processing, according to the following details .

DETERMINATION OF SOLIDS CONCENTRATION SINGLE FEATURE METHOD

A single feature can be used to calculate the solids content of a suspension. Fig. 6 shows an example of carbon in engine oil; it was measured at a temperature of approximately 84°C. This method is suitable for monitoring changes in solids concentration on an internal combustion engine (ICE) . A frequency of approximately 1.1 MHz minimizes the effects of particle size for this application. The particle size of the carbon particles created inside of an ICE are approximately 30 nm and larger. The size distribution of the carbon particles changes predictably as the smaller particles aggregate. Agglomeration of the aggregate particles also takes place as the carbon concentration increases. Fig. 6 shows three standing wave valleys showing frequency widths A, B and C, obtained during measurements of heavy-duty engine oils samples containing 1, 3 and 6% carbon respectively. The frequency width of the valleys is directly proportional to the solids concentration and is described by an almost linear relationship. Fitting this type of calibration data to a second order polynomial for multiple temperatures of interest allows the determination of solids content from signal voltage. The calculations required are familiar to those experienced in data interpretation and so need not be given herein.

REPETITIVE SCANNING METHOD

The most accurate solids concentration data is obtained from repetitively scanning the same standing wave peak. The Signal Interpret Algorithm 84 selects a peak and performs a frequency scan over the same frequency range, until the maximum frequency change has been found. The maximum frequency change f ax is found from: fmax = fl-fn; where fl is the modal frequency location of the first scan and fn is the subsequent scan that indicates the maximum frequency difference from f1. The test sample should not be flowing through the sensor 20 during the repetitive measurement period, since fluid flow minimizes the observed frequency shift. The fluid flow should be isolated by valve 48, shown in Fig. lb, during the entire measurement period, comprising one or more repeat frequency scans. Fig. 2 shows a "static fluid" embodiment of the sensor 20', suitable for accurate particle concentration measurements without fluid flow.

Fig. 7 shows the observed frequency shifts D and E for two test samples of used diesel engine oil that contain 0.6 and 4.1% by weight soot, respectively. The observed frequency shift varies linearly with soot concentration. A calibration procedure, comprising measurement of the frequency shift that occurs with test samples of known soot concentration, is familiar to those experienced in data interpretation. A linear fit is adequate for typical measurements in the range of 0.1 to 5%; and a 2^nd order polynomial regression fit encompasses a wider range, to more than 12% soot concentration by weight. The observed frequency shift fmax in soot contaminated engine oil is constant over a wide range of oil types and temperatures. The rate of frequency shift does however vary with sample temperature, fluid viscosity, fluid and particle density and other parameters, it is believed (but not confirmed) in a way familiar to those experienced in making particle mobility measurements .

DETERMINATION OF VISCOSITY

Fig. 8 shows a graph constructed from single standing wave peak measurements obtained for six mineral oil viscosity standards. These standards are available from CANNON Instrument Company of State College, PA. USA and Table 1 lists the data provided with these standards marked F - K in Fig. 8. Table 1 gives data at

25°C.

TABLE 1

Fig. 8 Standard Dynamic Kinematic Peak Name Viscosity, η Viscosity, μ Density, p

(cP)@ 25C (cSt)@ 25C (g/ml) @25C

S3 3.429 4.077 0.8411 S6 7.768 9.061 0.8573 H S20 29.02 33.93 0.8553

I S60 103.1 119.2 0.8648

J S200 413.4 469.6 0.8802

K S600 1368.0 1539.0 0.8888

The Signal Voltage (peak height) , when corrected for amplifier gain set on variable attenuator 78, provides a direct correlation with the test fluid dynamic viscosity η.

Fig. 9 is a graph of Signal Voltage versus Dynamic Viscosity constructed from the maximum signal voltage for each of the six mineral oil standard viscosity data sets shown in Fig. 8. The data in Fig. 8 and Fig. 9 were measured at a frequency equivalence of four wavelengths between the ultrasonic transducers. The dynamic viscosity values used in this plot were obtained from interpolation of the temperature / viscosity data provided by the manufacturer. Fig. 9 shows that the signal voltage (peak height) is inversely proportional to η. Dynamic Viscosity can be determined from an exponential fit to the data; an example is given in Fig. 9:

η = 4 _Λ2_r6_n7.5_r-e-2.3079X

where x is the Signal Voltage. The method just described provides a dynamic viscosity correlation for data measured at one temperature. However, it is advantageous to measure the changes in signal voltage that occur due to the variation of viscosity, since sensor temperature feedback on line 86 from temperature sensor 21 to microcomputer 62 in the system of Fig. 3 can affect the data.

Fig. 10 is a graph of signal voltage versus dynamic viscosity for Standard S60 mineral oil measured at multiple

temperatures between 25 and 40°C. The viscosity values plotted were obtained from the recorded temperature values from the sensor 21 and interpolation of the standard data given by the manufacturer. This method can be used to construct a series of viscosity / signal voltage data plots that cover the temperature range and dynamic viscosity of interest. Data fitting procedures familiar to those experienced in data interpretation are used to determine the relationship between dynamic viscosity and signal voltage at each temperature of interest. The recorded temperature and signal voltage then are used to determine the correct dynamic viscosity reading for subsequent sample test measurements . The method for performing this data fitting procedure is well known and is not detailed herein. DETERMINATION OF SOUND VELOCITY

The sound velocity through the sample fluid can be calculated from either a peak or a valley, or from the difference between two peaks or valleys. If the test sample contains particles, better results are obtained if a sharp valley is chosen to determine sound velocity, since minimal transfer of heat and movement of particles toward the receiving transducer occur during a frequency scan of a valley. The test sample characteristics should be approximately known, so that the wavelength equivalence of the feature chosen is known. For example, if the standing wave valley 92 in Fig. 5, corresponding to a wavelength multiple N of 2.75 and a transducer separation distance d, the sound velocity v can be calculated from:

= d/N . fmax (5)

where fmax_/ is the frequency of the modal value determined from

the standing wave data obtained. DETERMINATION OF DENSITY

Fig. 11 shows frequency scans of two standing wave peaks obtained in one "scan" for each of three mineral oil standards S20, S60 and S600 with data shown in Table 1. Calculating the sound velocity as above in Equation 5, using the modal frequencies measured for both peaks and assuming wavelength multiples of 3.5 and 4, yields different results for these three samples. The frequency differences shown as L, M and N in Fig. 11 correspond to 173,000, 176,500 and 179,500 Hz and are for standards S20, S60 and S600 respectively. These differences represent changes in sound velocity found between an even integer multiple number of wavelengths (4) and an odd fractional multiple of wavelengths (3.5). Since density of each fluid is constant throughout the scan, it can be inferred that the frequency differences L, M, and N include functions other than density. Plotting the locations of these two peaks for the whole series in Table 1 against each other, gives a linear plot with the known density values. This data gives the ability to monitor density within a known system. Plotting modal frequency for one standing wave valley versus one standing wave peak location does not however give a linear relationship. Correlating a standing wave valley with a standing wave peak involves phase velocity differences or wave shape informational. This non-linearity involves not understood, but might yield It is believed but not understood that values of the coefficient of compressional viscosity B, and kinematic viscosity can be determined from this data.

EXAMPLE OF EMULSION CONCENTRATION IN OIL

An emulsion is similar to a suspension in that one material is dispersed in another. Differences in the acoustic properties of any two materials with different acoustic properties allow concentration measurements to be made from sound velocity and attenuation measurements. Fig. 12 shows a plot of "scan" frequency versus signal voltage obtained for emulsion samples of water in soybean oil at the same temperature. The plots P, Q, R and S are of one standing wave peak and one valley and contain 0, 0.1, 0.5 and 1% by weight of water respectively. The data shows both frequency shift and signal voltage changes. The signal voltage changes at the valleys correspond to increased attenuation of the ultrasonic signal. The higher the signal attenuation, the less well defined are the valleys. Although the data is not reduced to concentration measurements herein, the differences given in Fig. 12 show that useful concentration data can be obtained assuming constant particle size data. It is believed that the herein above significant presentation of this invention will enable those skilled in the art to employ the method and devices technology disclosed herein, with suitable modification where needed, without departing from the spirit and scope of the invention as claimed.

Claims

WHAT I CLAIM IS :

1. A standing wave interferometry analysis method for characterizing the physical properties of a test sample of at least one of: fluid, suspension, emulsion; said method comprising the steps of: a) defining a test sample sensing zone to be in the proximity of at least one ultrasonic transducer; b) placing test sample into the sensing zone; c) applying to the transducer and the sensing zone an ultrasonic frequency; d) changing continuously, by small intervals, the ultrasonic frequency, thereby; e) creating at least one standing wave feature in the sensing zone; and f) analyzing the standing wave feature for at least one of: amplitude, frequency location, frequency width.

2. The method according to claim 1 and the step of: obtaining test sample characterizing data from said step of analyzing.

3. The method according to claim 2 and the step of: calculating, from the obtained data, at least one of: particle concentration, viscosity, density, sound velocity of the test sample.

4. The method according to claim 1 in which; said step of defining is between two ultrasonic transducers .

5. The method according to claim 4 and the step of: spacing the transducers part by a distance equal to wave length multiples.

6. The method according to claim 1 and the step of: operating the transducer in its near field region.

7. The method according to claim 1 in which said step of creating provides peak as one feature.

8. The method according to claim 1 and the step of: scanning the same standing wave feature repetitively.

9. The method according to claim 8 and the step of: measuring the resultant frequency shift.

10. The method according to claim 1 and the step of: utilizing a test sample having sub-micrometer particles.

11. The method according to claim 1 and the step of: utilizing a test sample of engine oil having therein a concentration of carbon.

12. The method according to claim 1 and the step of: measuring the temperature of the sensing zone at each frequency interval .

13. A sensor for use in standing wave interferometry analysis of a test sample, said sensor comprising: a) a pair of spaced apart ultrasonic transducers; b) a test sample sensing zone defined in part by the space between said transducers; c) said space between said transducers enables them to operate in their near field region; d) the distance said transducers are spaced apart being wave length multiples of a standing wave; e) one of said transducers being arranged and connected as a signal transmitter; and f) the other of said transducers being arranged and connected as a signal receiver.

14. A sensor according to claim 13, further including; a temperature sensor positioned to be responsive to the temperature proximate to said sensing zone.

15. A sensor according to claim 13, further including: a test sample inlet and an outlet, positioned such that test sample passing therebetween will pass smoothly through said sensing zone.

16. A sensor according to claim 13 which is constructed and arranged to respond to small intervals of change in ultrasonic frequency.

17. A sensor according to claim 16 and in combination therewith; means for generating an ultrasonic frequency which is continuously changing by small intervals; and means for applying said changing ultrasonic frequency to said sensing zone.

18. A sensor according to claim 17 and in combination therewith; electrical wave shaping means coupled to receive from said signal receiver standing wave features which characterize physical properties of test sample.

19. A sensor according to claim 18 and in combination therewith; interpret algorithm means for quantifying said standing wave features after they have been processed by said electrical wave shaping means; said quantifying providing information concerning the test sample, including at least one of: viscosity, density, sound velocity, particle concentration.

20. A sensor according to claim 13 which is constructed and arranged to analyze a test sample containing sub-micrometer particles.