"Development of a sonar for finding gas compositions" J. C. Vyas, Crystal Technology Section, Technical Physics Division, BARC.

"Development of a sonar for finding gas compositions"

J. C. Vyas, Crystal Technology Section, Technical Physics Division,

BARC

Out line

• Introduction

• Gas sensors

• Energy transfer in gases

• Our system

• Some results

• Conclusion

Introduction

• Usually, gases are identified by more elaborate physico-chemical methods, such Chromatography, or by Mass spectro-meters, or by IR absorption line intensities etc.

• Most of these techniques need elaborated procedures and time to find or identify the composition of gas mixtures.

• Gas handling and utility applications need quick and rugged methods to find gas composition.

A gas sensor• An ideal gas sensor is that which can find (identify) a

specific gas quantitatively, in a mixture of two or more gases, in shortest possible time, and without damaging the composition of the gases. The system should be – Rugged– Able to detect any gas in a given gas mixture– Possibly non-invasive– Smallest response and dead time (if any)– With best possible resolution – Stable (with respect to time and constituent components) – Without (or minimum of) the wear and tear or aging problems, and – Easy to handle, etc

Types of gas sensors

• No ideal gas sensor system exists at present commercially, but with compromises on some of the requirements, solutions can be found. There are several types of gas sensors available, such as– Infrared absorption based

– Chromatography based

– Thin film (redox-conductivity) based solid state

– Fibre-optic based solid state

– Mass spectroscopy based, etc

• We searched in the ultrasonic field and fabricated an Ultrasonic Gas Detection System (a kind of gas SONAR).

The Sonar

• A sonar is acronym for (a device that uses) sound energy pulses for navigation and ranging.

• The device is used in several non-destructive methods, such as depth finding in deep waters (sea / river beds), or mechanical strength of structure. A modified version is used for imaging (medical use) of human organs (such as in sonography), etc.

• Can such technique be used for measuring gas composition ?

Energy transfer• In materials, mechanical energy transfer takes place via

physical collisions. For elastic collisions both energy and momentum need to be conserved.

m1u1 + m2u2 = m1v1 + m2v2

• And (1/2){m1u12+ m2u2

2 = m1v12 + m2v2

2}

• If the body is rigid, the mechanical energy transfer within the body material is via conduction mode, but in fluids an additional convective mode also come into play in some proportion.

• We neglect direct radiation part (in the low energy per particle regime).

Speed of energy transfer

• The speed of mechanical energy transfer within specific material is the speed by which the energy travels in a material from one site to another, and known as speed of sound in common terminology.

• This speed depends upon the response time of nearest neighbour, which in turn related with the rigidity or the strength of bonding within the structure. So in solids the speed is relatively larger, and largest for the strongest (bond) material. In diamond it’s value is 17500 m/s and reflects further in large thermal conductivity (at R.T. it is 20-25W/cmK, exceeds that of copper by a factor of five).

Energy transfer In fluids

• Fluids differ from solids. In case of liquids we have a fixed volume during mechanical energy transfer, but flexible movement of constituent molecules.

• Gases are a part of fluids, but volume has a strong relation with the pressure term. The volume and the shape depend upon the container vessel.

• In gases the physical parameters of basic concern are the gas pressure, and the temperature of the gas medium.

Sound speed in Gas media• Gases are special fluids, in which

sound / ultrasound velocity depends upon externally controllable physical parameters, the pressure, the specific heat ratio, and the average molecular weight (density) of the medium.

• If the physical parameters of the gas medium can be controlled by external means properly (to remain confined around certain fixed value), the sound / ultrasound velocity becomes a direct function of gas density.

Some formulation

• In gaseous medium for a reasonably constant temperature

conditions, the sound (or ultrasound frequencies 10 MHz)

speed is given by

v = ( P / )0.5

Here, v = Speed of sound / ultrasound (m.s-1)

= Ratio of molar heat capacity of the gas (Cp/Cv)

P = Pressure of the gas medium (in Pa)

= Density of the gas (kg.m-3)

What happens in gas mixtures• For a gas mixture this relation becomes (for constant

temperature conditions)

vmix = (mixP / mix )0.5

vmix = Speed of sound / ultrasound for gas mixture (m.s-1)

mix = Ratio of molar heat capacity of the gas (Cp/Cv)mix

P = Pressure of the gas medium (in Pa) for mixture

mix = Density of the gas mixture (kg.m-3)

• The sound speed changes with the density of gas mixture, and

may be used to provide the composition of unknown gas in the

mixture, provided the mix is also known for the gas mixture.

The next job: find relations• The density of a gas mixture for densities of individual

gases ρi and gas concentration xi for the ith species (with normalization xi = 1), may be given by

ρm = xi ρi

This relation for only two gases with x = x1 and so x2 = (1-x) becomes

ρm = xρ1 + (1-x)ρ2

• Similarly, the relation for molar specific heat ratio of a gas mixture be written. For the case of two gases only can be expressed in terms of individual molar specific heat capacities in first approximation, with as above notations

• m = { xi Cpi / xi Cvi }={xCp1 + (1-x)Cp2}/{xCv1 + (1-x)Cv2}

Strategy

• Fabricate a system, which can measure the speed of sound for any gas / gas mixture. The system should be handy, and must have capacity to provide data as fast as possible (almost on line), and look at the results.

Practical problems • Sound speed in gas media is independent of frequency from ultra

low to few MHz (for constant P and T conditions).

• Sound from the mechanical sources in the lab or industry, may complicate measurements (up to about ~ 40kHz, and so go to higher frequencies).

• Higher frequency (a better precision, an advantage!), in the measurement of sound speed, and this also reduces the size of the measurement cell.

• However, higher frequencies also have some basic problems – Attenuation (increases with increasing frequency in exponential way) – Fabrication of suitable transducers becomes more difficult

The scheme• Armed with the functional relations of the density, and the

specific heat ratio of the gas mixture with corresponding individual gas parameters, measurement of sound speed in the gas mixture, and in one of the individual gas (the background gas), should provide the unknown concentration of specific gas in the gas mixture

• The information on the gas density, and the molar specific heat capacity ratio, can be obtained from the literature or elsewhere or by direct measurements.

• For measuring the sound speed, a small pulse be sent through the gas media over a fixed distance and it’s eco be registered in time (known as time of flight (TOF)).

Control of other parameters• Gas pressure and its temperature are main

parameters, to be controlled properly.

• In normal ambient air, the pressure is nearly one atm., (= 105 Pa), and relatively easier to control.

• For flowing gas conditions (at normal flow), the pressure difference may be controlled. Further, in the speed expression we have (P / ), which is almost a constant quantity, at normal pressures

• The temperature of the gas media is certainly a more difficult parameter to control with an arbitrary accuracy. However, it can be controlled to some extent by use of thermal baths etc., but still with certain level of inherent fluctuations.

Process of Measurement

• In practice, for all the signals traveling to a fixed distance, one can use the

TOF values directly, to find out gas concentration in the sample.

• In pulse catch mode one needs two transducers operating at same frequency.

However, one can use a single transducer and register the reflected (echo)

signal in time.

• An ultrasound signal (~500 kHz) of short duration (signal width about 5 to

10 ) is created and passed through the gas medium

• The reflected pulse after traveling a path-length of about 500 , is detected

and the TOF is recorded

Process of … …

• The Time of Flight (TOF), is measured for the gas mixture tm, and for the

pure background gas t1 and t2 , and the gas concentration x is calculated

(for low concentrations of the unknown gas) by

x = (tm - t2)

With = (tm+t2)/(t12-t2

2), and t1 is the TOF for pure unknown gas.

It is possible to measure graphically also for low concentrations and so

measurement of parameter t1 is not always needed.

• For small concentrations of unknown gas, parameter , may be

considered as an approximate constant for the cases tm t2.

The electronics and measurement cell

• We have used a commercial pulser-receiver model 4400 MX, which can measure TOF with resolution of about 20 ns in the 0 to 2 ms range.

• The transducer generates ultrasound at central frequency 500 kHz in pulse mode, and also detects the echo of the reflected signal.

• The transducer is fixed in a closed test chamber (vol. ~60 cm3) filled with normal air at 1 atm., pressure and it has provision for insertion of sample gas, into it.

Experiments• Temperature of the test chamber was maintained at 25 or 260C, using a water

jacket.

• The pulser sends a short signal (pulse) of about 20 s for every 20 to 50 ms time interval, and the same transducer detects the reflected eco. In present cell design TOF of about 700 - 850 s (for the background gas) measured.

• TOF values were recorded for samples of clean air with time, and average value was taken out (the value of t2), it was 793 s. It was seen that within lab

the variation in t2 was less than 0.1 s for a time period of about 30 min.

• A known amount of gas say SF6 was injected into the chamber (cell) using

calibrated syringe and corresponding TOF data was recorded. This step was repeated for several different concentrations of the test gas. A typical set of data based on observed TOF v/s SF6 gas concentration is shown in next frame.

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60792

793

794

795

796

797

798

799

800

801

802

803

804T

OF

va

lue

s (

mic

rose

c.)

SF6 Gas Conc. (mol %) in Air

Resolution• Electronically we resolve down to 20 ns. The slope of line in graph,

indicates that 0.0559 vol % (~560 ppm) of SF6 gas in air, corresponds to 1

s change in TOF, and so, ideally (in present set up), we can resolve 10 ppm of SF6 gas in air!

• According to our own study the temperature fluctuations T cause variation in TOF values, say t, and the correspondence is given by (T/T) = 2(t/t)

• For a typical case, assuming T (± 0.5K), at T = 300K, corresponding t value is (t/1200). Let a typical value of TOF, is t ~ 800 s, for this case the spread t comes out to be 0.67 s, which corresponds to the SF6 gas

resolution of about 400 ppm in air. • Notice, that if we can maintain T ± 0.1K, the corresponding resolution

for SF6 gas becomes around 100 ppm. But to get best resolution by this set

up, that is 10 ppm, the temp. fluctuations must be limited to T ± 0.01 K.

Response time

• The system does not have a measurable dead time. However, the pulse repeating period may be considered as the time in which next set of data on TOF will be available, this time is about 20 to 50 ms.

• The time in which the injected gas molecules diffuse in the measurement cell is of few millisecond (ms) order, and can be considered negligible.

• Since, in 1 sec we have about 20 to 50 data points, the system is capable to work as almost online for detecting sample gases.

• It may be noted that gases do take a finite time to reach from the leakage site to the sensor location. This part varies with respect to different site locations, the relative position of the sensor w.r.t. source position, and therefore can not be taken into account.

Limitations of the system (important !)

• This system measures a change in the average density. So if one is looking for possible leakage of a specific gas, he is not always sure in case if it comes from some other gases present in the lab.

• However, such limitation is in fact, a blessing in disguise for the safety of equipments and personals, as one gets information on change in gas ambient immediately, and may provide an early alert to control system (if needed) about such an event of other gas leak.

• The gas composition resolution depends upon the nature of sample gas as well as the background gas parting in the relative density of gas mixture, and so it’s value will be different for different combinations.

Conclusions

• Ultrasonic pulse sensor can be used to detect gas compositions or gas leakage almost on line (more than 20 data points per second can be obtained in present set up).

• In present set up although ultimate resolution (electronically possible) one can get is about 10 ppm, for SF6 gas, but actual gas resolution is limited by the temperature fluctuations of the gas media in the cell.

References

• A. B. Bhatiya,“Ultrasonic Absorption, An introduction to the theory of sound absorption and dispersion in gases

liquids and solids”, Oxford, Clarendon press, 1967

• G. Hallewell, et al, Nucl. Instr. Meth. in Phys. Res., A264 (1988) 219

• Vyas J.C. et al, “A non-invasive Quantitative Method for H2 Gas Detection in Air,” DAE Solid State Physics

Symp. 2004, p-342

• J.C. Vyas et al, “A non-invasive ultrasonic gas sensor for binary gas mixtures,” Sensors and Actuators B 115

(2006) 28

• J.C. Vyas and V.R. Katti, “Effect of temperature fluctuations on resolution of ultrasonic non-invasive gas sensor

for gas/air binary systems,” Proc. of 12 th Natl, seminar on phys and Techl., of sensors (NSPTS-12, march 7-9,

2007) ed: A.K. Debnath and S.K. Gupta, p-158

• J.C. Vyas at el, “Sensing of SF6 gas in air using non-invasive ultrasonic gas sensor,” International Conf., on

sensors and related networks (SENNET –07) Dec 12-14, 2007, VIT Univ, Vellore (Tamilnadu) India, p-82

• J.C. Vyas, and V. R. Katti, “Ultrasonic gas sensor as secondary standard for composition measurement of binary

gas mixtures, DAE solid state physics Symp. Myssore Univ., Myssore,” Ed: Amitabh Das et al, 52 (2007), p-413

• J.C. Vyas at el, “Sensing of SO2 gas in binary mixture with N2 gas using ultrasonic gas sensor and its comparison

with other standard methods. Exploration and Research for Atomic Minerals,” Vol. 19 (2009) 233-237.