3000
IEEE SENSORS JOURNAL, VOL. 16, NO. 9, MAY 1, 2016
Rapid and Automated Measurement of Milk Adulteration Using a 3D Printed Optofluidic Microviscometer (OMV) Pedinti Sankaran Venkateswaran, Abhishek Sharma, Santosh Dubey, Ajay Agarwal, Senior Member, IEEE, and Sanket Goel
Abstract— In developing countries like India, adulteration in the milk consumed by the population presents stern implications as tarnishing of the same poses serious issues, such as health deterioration, corruption, and so on. There are many adulterants that are added to milk, including water, flour, starch, and even urea, in quantitative measures making it undetectable. There are many devices in the market to measure adulteration in milk but most of them are bulky, require large sample volume, and need a technical operator for working. In the recent decade, microfluidics has emerged as a huge market for biomedical research. It has paved the pathway for a quick, robust, and plug-and-play device for various applications. This paper describes a low-cost, durable, and simple optofluidic microviscometer fabricated by the stereolithography technique. The device operation is based on the linear relationship between dynamic viscosity and channel width derived from the flow of two immiscible fluids inside a channel. The principle of operation is based on the modified Hagen–Poiseuille flow equation. The working principle is the viscosity-dependent capture of the microchannel width by the fluids flowing inside the microchannel under the laminar flow based on the pressure gradient between the inlets and the outlet. In this paper, around 60 milk samples with various adulteration ratios of various adulterants ranging from 1% to 10% have been tested. A best fit curve for every adulterant was defined, and the device was found to be accurate enough to measure the entire range of adulteration ratios with a high accuracy of 0.95. Index Terms— 3D printing, microviscometer, milk adulteration, dynamic viscosity, laminar flow, microfluidics. Manuscript received January 4, 2016; revised February 5, 2016; accepted February 7, 2016. Date of publication February 11, 2016; date of current version March 16, 2016. This work was supported by the Department of Science and Technology, Government of India, under the Technology Systems Development (TSD) Scheme for this project (#DST/TSG/ME/2012/08). The associate editor coordinating the review of this paper and approving it for publication was Dr. Anna G. Mignani. P. S. Venkateswaran is with Research and Development, Ark Infosolutions pvt. Ltd., Bengaluru 560011, India, and also with Research and Development, University of Petroleum and Energy Studies, Uttarakhand 248007, India (e-mail:
[email protected]). A. Sharma is with the Department of Electronics and Instrumentation Engineering, University of Petroleum and Energy Studies, Dehradun 248007, India (e-mail:
[email protected]). S. Dubey is with the Department of Physics, University of Petroleum and Energy Studies, Dehradun 248007, India (e-mail:
[email protected]). A. Agarwal is with the Council of Scientific and Industrial Research, Central Electronics Engineering Research Institute, Pilani 333031, India (e-mail:
[email protected]). S. Goel is with the Department of Electrical and Electronics Engineering, Birla Institute of Technology and Science–Pilani, Hyderabad 500078, India (e-mail:
[email protected]). Digital Object Identifier 10.1109/JSEN.2016.2527921
I. I NTRODUCTION ILK IS the next highest consumed liquid commodity after water in the world. But to some, it is a trade which reaps rich dividends, particularly in developing countries, like India [1]. In India, majority of the milk supply depends on the individual units set up by the dairy farmers of the states as it is important for sustaining the livelihoods of many smallholder farmers who contribute to nearly 70% of the total milk production in the country [2]. They go about supplying the milk to the bigger factory setups for processing and final distribution. This makes them a part in the entire process chain of milk production and distribution. In this case the profit margin scored by the individual milk producer remains constant and increases marginally state-wise based on various parameters as put forward by the Department of Animal Husbandry, Dairying & Fisheries Ministry of AgricultureGovernment of India [3]. Hence, these individual producers tend to go by distributing their produce on their own. This is where the pervert problem of adulteration exists [4]–[6]. Most of these milk suppliers tend to play around with the viscosity of the milk that they supply in order to increase the quantity and thereby reap profits in a small scale manner. This small scale corruption results in a chain reaction when we take into account the mass of such milkmen reaping such benefits. The adulterants range from water, starch, corn-flour to urea and various detergents. Though the adulteration scale is small, they reap better paybacks compared to that provided by companies at the cost of the consumer’s health and safety [7], [8]. Hence, there is a need of a device to measure the adulteration of milk real-time. Amongst the various available devices to test the purity of milk, the major problem exists with the size, accuracy and cost of the device. Historically, various tests that can be performed to test the purity and quality of milk include organoleptic test, clot on boiling test, alcohol test, alcoholalizarin test, acidity test, resazurin test and the Gerber Butterfat test, ac conductance test [9]–[12]. All these tests require complex measurement, costly equipment, lot of time and an experience operator. The major hurdle in all of them that none of the test are user friendly i.e. a common man cannot use them on a daily basis to test the milk that they receive. Besides, the parameters used to test are more complex for basic understanding. But, if one can carefully calculate and
M
1558-1748 © 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
VENKATESWARAN et al.: RAPID AND AUTOMATED MEASUREMENT OF MILK ADULTERATION
do some basic analysis, it becomes simplified to determine that viscosity forms one of the most important parameter in the testing of milk adulteration [13]. Viscosity is the rheological property of a fluid that can be demarcated as the resistance to fluid flow [14], [15]. There are a few standard laboratory viscometers which are used to measure the viscosity of fluids. Most commonly used viscometers are U-shape glass viscometer, Rotational and Vibrational viscometer [16]. All these viscometers necessitate large sample need, controlled environment and expensive equipment [17]. A real-time monitoring of the viscosity variations w.r.t to adulteration is still an emerging field of research. Microfluidics has given the world with a variety of technologies for a wide range of applications from chemistry to biology [18]. Miniaturized sensors and systems developed by this technology route can be used for a variety of applications [19]–[21]. Microfluidic devices offer numerous advantages over their macro sized counter-parts like lesser sample need, quicker analysis, real-time measurement and high accuracy [22]–[24]. Some work has been carried out to conceptualize, design, fabricate and test a microviscometer. A range of materials, such as silicon [25], glass and different polymers (SU-8 and PDMS) [26], [27] have been used to realize it. Chevalier and Ayela [28] presented a micromachined capillary on chip rheometer using anodically bonded silicon-Pyrex derivative microchannels equipped with local probes, and used silicon oil and ethanol-based nanofluids. With this device, the local pressure drop can be measured inside the microchannels di-electrophorecially without the need of reference fluid. Chevalier and Ayela [28], Zuoyan and Zheng [26], and Zuoyan et al. [26] demonstrated a polydimethylsiloxane (PDMS) micro-fluidic device for measuring the viscosity of Newtonian fluids by using the high solubility and permeability of air in PDMS to generate vacuum (and get pressure differentials) in the degassed PDMS micro- fluidic device. This device was used for different types of samples, such as, glycerol, proteins, blood plasma and organic solvents. However, these devices were found to be complex and were majorly laboratory based ones. This made the possibility of realizing a field device an open area of research. The prime aim of this paper is the design, fabrication and experimental analysis of a simple optofluidic microviscometer that can measure the variation in the dynamic viscosity using the modified Hagen-Poiseuille flow equation of an unknown sample when it is made to flow alongside an immiscible fluid of a known viscosity. The device records the interface position of the adulterated milk sample and the immiscible reference fluid glycerin in a common channel. Further, the device which has been designed by Goel and his team of researchers can also be used to monitor automobile fuel adulteration and biodiesel blending wherein the reference fluid can be made to flow with the unknown quantity of fuel and the dynamic viscosity of the unknown fluid can be calculated [29]. The microviscometer requires the aid of an optical microscope to measure the changes in the width occupied. However, this can be replaced by a small commercial microscope which can significantly reduce the cost and ease of operation. This microviscometer offers many advantages like better accuracy, lower cost, real-
3001
Fig. 1. Flow of two immiscible fluids between a pair of horizontal plates under the influence of a pressure gradient.
time measurement, portable and easily operated device compared to the commercially available counter-parts [30]–[34]. The meek and handy device design makes it compatible for many other applications as well. The adulteration of milk with various commonly available adulterants at different ratios of adulteration was successfully tested in the laboratory and the results have been explained with detailed analysis in the sections below. II. M ATERIALS AND M ETHODS A. Flow of Two Adjacent Immiscible Fluids Inside a Microchannel The investigation of fluid flow in micro channels has attracted great interest in the last decade because of the variety of potential biochemical applications [35]. Many researchers and investigators have used the Navier–Stokes equations to describe the laminar flow in micro-tubes and micro-channels [36]. Based on the derivation performed by Bird et al. [37] for the flow of two immiscible fluids in a rectangular channel, we have also considered two immiscible fluids flowing in a rectangular channel of channel of length L and width W under the influence of a horizontal pressure L gradient P0 −P L . Our derivation, however, is different than the one performed by Bird et al in the sense that we have fixed the flow rates of the fluids at the input and estimate the width occupied by both the fluids as they flow along the channel, whereas Bird et al adjusted the flow rates at the input such that the width occupied by both the fluids is same in the channel. The velocity profile of both the immiscible fluids in a rectangular channel has been displayed schematically in Fig. 1. In this figure, we see that both the fluids have different velocity profiles across the interface; the velocity at the interface is the same. The width occupied by both the fluids has been also different, which we found experimentally. The momentum-flux for the two fluids flowing in and occupying specific regions in the channel can be written as P0 − PL (1) x + C1I τxIz = L P0 − PL (2) τxIzI = x + C1I I L The momentum flux is considered to be continuous throughout the fluid-fluid interface; hence applying this as a boundary condition, we get: B.C. 1: at x = 0, τxIz = τxIzI
3002
IEEE SENSORS JOURNAL, VOL. 16, NO. 9, MAY 1, 2016
Therefore C1I = C1I I = C1 . Equations 1 & 2 become: I P0 − PL I dv z −μ = x + C1 dx L dv I I P0 − PL −μ I I z = x + C1 dx L
(3) (4)
The above two equations can be integrated to obtain the velocity equations for the two fluids: P0 − PL C1 I (5) x 2 − I x + C2I vz = − I 2μ L μ P0 − PL C1 (6) v zI I = − x 2 − I I x + C2I I 2μ I I L μ
Fig. 2.
Geometry (in mm) of the optofluidic microviscometer.
In the case of two immiscible fluids flowing in the same channel, the flow rates can be written as
Using the fact that the velocity of both the fluids at the interface is equal and the conditions for no-slip at both the boundaries along x-direction, the integration constants may be estimated: B.C. 2: at x = 0, B.C. 3: at x = b1 ,
v zI
=
v zI
v zI I
QI I
B.C. 4: at x = −b2 , v zI I = 0 Applying the boundary conditions in equations 5 & 6, and substituting the values of the constants, the resulting velocity profiles would be: II 2 I b2 μ b − μ P − P − P P x 0 L 0 L 1 2 x2− I v zI = − 2μ I L μ 2L μ I I b1 + μ I b2 b1 b2 (b1 + b2 ) P0 − PL (7) + 2L μ I I b1 + μ I b2 μ I I b12 −μ I b22 P0 − PL P0 − PL 2 x II vz = − x − 2μ I I L μI I 2L μ I I b1 + μ I b2 P0 − PL b1 b2 (b1 + b2 ) + (8) 2L μ I I b1 + μ I b2
(9) (10)
2 II d 2 v zI (P0 − PL ) I I d vz = μI = μ (11) L dx2 dx2 According to Hagen-Poiseuille’s law for laminar flow in a rectangular channel [38]:
Pbh 3 12Lμ
Here P is the pressure difference between inlet and outlet b is the width occupied by the fluid in the channel h is the height of the channel μ is the viscosity of the fluid L is the length of the channel
(14)
P I b1 P I I b2 = I L μ L μI I
(15)
We know that d 2vz P (16) =μ 2 L dx Replacing P/L in equation 15 with the help of equation 16 we get: μI
2 II d 2 v zI b1 b2 I I d vz = μ 2 I 2 dx μ d x μI I
Using equation 11, we get b1 b2 = II μI μ which may be rearranged as
From both the equations, we see that:
Q=
(13)
If the flow rates are adjusted to be the same for both the fluids, then
=0
Upon double differentiation we get d 2 v zI P0 − PL = − dx2 μI L 2 I I d vz P0 − PL =− dx2 μI I L
P I b1 h 3 12μ I L P I I b2 h 3 = 12μ I I L
QI =
(12)
b1 μI = I I μ b2
(17)
Therefore, the viscosities of two immiscible fluids flowing inside a horizontal channel (of height h) and occupying widths b1 and b2 respectively, will be related by equation 17. B. Design and 3D Printing (Fabrication) of Y-Shaped Microviscometer The interface shift analysis was carried out in a 3D printed optofluidic microviscometer with a y-shaped channel. The entire device was designed using Rhinoceros software (Version 5.10). As shown in the Fig. 2, the device design consisted of repetitive loops of channels for the controlled laminar flow inside the channel [39]. The flow rate at the inlet was approximately set as 4.5 μL/min to maintain the same pressure differential. The 3D printer used to print this device was MiiCraft Kit (PN#95.LF800G004) from the Rays Optics, Taiwan. This is a stereo-lithography (SLA) based printing machine with a minimum resolution of 56 microns across the XY axis and
VENKATESWARAN et al.: RAPID AND AUTOMATED MEASUREMENT OF MILK ADULTERATION
Fig. 3.
3D printed output of the optofluidic microviscometer.
50 microns across the Z axis. It uses the bottom-up approach of printing based on resin curing technique [40]. The output file from the design software should essentially be saved in the STL format for the MiiCraft STL Viewer to position the design on the platform of the printer. This is followed by the Slicer program which slices the entire design into 50/100 micron individual slices and generates the final index file for printing in the ‘Print STL Model’ module. The printer usually takes less than 5 minutes to print a model of height 1 mm using a standard available UV curable polymer input material. And after the printing it usually goes into a 5 minute post-curing cycle for the final product as shown in Fig. 3. The input material used in a transparent resin (MA-YG2005T) supplied by the company. It is a transparent liquid having a flash point of 150°C and a specific gravity of 1 gm/ml at 25°C. In a typical printed device, the dimensions of the channels were kept at 1 mm overall (in X, Y and Z planes) and the diameter of the inlet and outlet ports were 2.6 mm wherein the tubes for the fluid flow inside the channel would be fixed. The thickness of the channels was also kept as 1 mm. The microchannels were rectangular in nature and printed using the layer by layer printing fashion of the SLA printer. The length and breadth of the device was fixed to 4 cm × 2.5 cm as the maximum printable area of the 3D printer was limited. Besides, for the analysis and measurement, this dimension were more than sufficient as the number of loops extended helped in the proper laminar flow in the channel. The length of the over lopped channels were chosen to be as large as possible in the printable area of the printer. This was done in order to make the flow of the liquids to be laminar in nature. The length of the main channel was kept as 27.5 mm as shown in the Fig 3. This was more than the required length of 15 mm which was calculated based on extensive measurements. The design was aimed to maintain a laminar flow in the field when the two streams, A and B, are united and thus prevent uncontrolled convective mixing [41]. The interface position for measurement was taken at the exact midpoint of the length of the main channel i.e. 13.75 mm. Various angles of approach were studied ranging from 5 degrees to 90 degrees. The higher the angle of approach, the more the instability in flow was found. Hence, after careful deliberations and testing, the best approach angle for each side of the channel for the analysis was found to be 10 degrees. C. Sample Preparation-Commonly Used Adulterants in Milk The samples that were tested using the 3D printed microfluidic viscometer were prepared based on the careful study
3003
and analysis done by students in the local community of the various states in India. The individual milk producers and distributers do not share such critical information as it may lead to a total closure of their product selling. Hence, careful tactics were applied wherein close bonding with these producers were developed thereby leading to the discovery of the various adulterants used in the milk being sold locally. The major adulterants were found as water, flour (corn and wheat based), starch (corn and potato based), urea and detergents. The use of detergent was just to produce the soap bubble effect in the increased quantity of milk (after the addition of water) and hence the quantity was found to be very miniscule to be considered for testing [42]. In the case of urea, it is used in a very minute quantity for the enhancement of the viscosity in the milk so that water can be added to further increase the quantity. The major adulterants as per the usage statistics are water, flour and starch. The milk sample taken for the investigation was from the local distributer after careful bargaining and tested using the conventional method (Anton Parr Rheometer). The average dynamic viscosity of the pure milk at 20°C was 19.10 mPa.s. The reference fluid glycerol was procured from Sigma Aldrich India and the average dynamic viscosity was found to be 430.70 mPa.s at 20°C. The viscosity of the unadulterated and adulterated milk can also vary depending on composition, temperature and the age of the milk. But the variation w.r.t to these was not considered for significant for the analysis. The concentration of each adulterant in milk was judged and prepared based on the direct inputs from these milk producers. In the case of flour, the usual mixing was in the range of 150 to 200 grams in 5 kilograms of milk which would amount to 3 to 4 wt. % of the overall concentration. Starch was usually taken as 100 to 150 grams in 5 kilograms of milk resulting in 2 to 3 wt. % of the adulterant in milk. Since urea was found to be toxic, only a meagre 1 to 2% was added in milk. The samples with adulteration were prepared from 0.5% till 10% for the solid adulterants and water was added to this ratio only in the case of flour. In the case of adulteration of milk with water, ranges with 5-95% water was prepared as the minimum quantity for accurate measurement was found to be 5%. Hence, in total 80 samples were prepared and tested using the optofluidic microviscometer and Anton Parr rheometer for the conventional measurement of viscosity. D. Experimental Setup The experimental setup comprised of the 3D printed optofluidic microviscometer, two peristaltic pumps for inlets, silicone rubber tubing’s, pure glycerol as the reference fluid, adulterated milk samples and a Leica Microscope (Model-DFC3000G) for width occupancy measurement and analysis. The entire schematic of the setup was arranged in a fashion as shown in Fig. 4. The reference fluid was taken as pure glycerol and the samples of the adulterated milk were taken as mentioned in the previous sections for the testing and analysis. All the experimental investigations were carried out at standard atmospheric pressure of 1 atm and temperature of 20° Celsius. All the experiments were carried out inside the laboratory considering
3004
IEEE SENSORS JOURNAL, VOL. 16, NO. 9, MAY 1, 2016
Fig. 4.
Schematic of the experimental setup.
Fig. 5.
Fig. 6. Biphasic interface position measurement for flour in milk using the OMV.
Actual eperimental setup.
all the precautionary measures and safety standards as shown in Fig. 5. As per the Hagen Poiseuille flow equation, the flow rates of two immiscible fluids may be adjusted so that they occupy the same width inside the channel. However, if the flow rates are fixed, then the liquids flowing in would occupy widths based on their viscosity-dependent velocity. The two liquids in the OMV were made to flow inside the channel at the same flow rate of 4.5 μL/min. This is as per the real-time experimentation for a rectangular channel. The flow rate can be changed in a linear fashion if the pressure differential across the channel needs to be increased or decreased. The measurement point of the width occupancy was chosen to be one-third of the total length of the common channel i.e. 9 mm. The fixing of this to be the measurement point was based on the fact that the two fluids come into contact at the start of the common y-channel and after a few mm both the fluids start to flow in a smooth fashion inside the microchannel occupying a said width based on their viscosities. The Fig. 6 shows the measurement of the sample fluid inside the microchannel using the live recording of the fluorescence microscope for flour adulteration in milk at 2%, 4%, 6% and 8%. III. R ESULTS AND D ISCUSSIONS All the adulterated samples were tested using a conventional Anton Parr Rheometer (Model- C-LTD80/QC) for their dynamic viscosity. The rheometer though has a fairly high rate of accuracy but the time taken by it to give an accurate reading of the dynamic viscosity was in the range of 10 to 15 minutes (average 200 readings). The rheometer was operated at 1000 rpm (revolutions per minute) measurement cycles. The higher rpms are generally used for less viscous
Fig. 7. Width occupied by water adulterated milk sample vs sample viscosity and % of adulteration.
samples and the lower rpms are used for highly viscous samples. In the present study most of the samples were of the less viscous nature. The various adulterant based test results are discussed in the further sections. A. Water Adulteration in Milk The Fig. 7 clearly shows the variation of channel occupancy width with the decrease in the viscosity of the milk due to the gradual increase in the water content in the milk. Water is the most commonly available adulterant for milk whereby the adulteration quantity can even be to an extent of 75% [43]. Upon careful observation, it is evident that the regression plot is of a polynomial nature. The reason could be attributed to the fact that water is miscible in glycerin. Hence during the flow of the two liquids in the common channel, there could have been a possible diffusion of water from the milk to glycerine. We also see that the linear part of the fit has larger coefficient (0.0523) than the quadratic part (0.0009). So, linear terms dominates over the quadratic term. Despite this, the case of a single solution (despite both being different fluids initially) inside a channel for which the fluid follows a parabolic behavior. However, Sahu et. al have also proved that two solutions of the same cadre with varying viscosities can also lead to the development of a line of a demarcation based on modified flow rates and weak level of diffusion [44]. Hence, further analysis on the flow pattern and flow rate can result in the deduction of a possible methodology to measure
VENKATESWARAN et al.: RAPID AND AUTOMATED MEASUREMENT OF MILK ADULTERATION
Fig. 8. Width occupied by flour adulterated milk sample vs sample viscosity and % of adulteration.
3005
Fig. 10. Width occupied by starch adulterated milk sample vs sample viscosity and % of adulteration.
B. Flour Adulteration in Milk The most commonly used adulterant in milk is flour [45] as the mixture with hot water remains almost undetectable. This is mixed in hot water before it is made to mix with milk. The concentrations of flour used in this experiment are based on equal additions of hot water in them and then mixing it with milk. The linear regression graph showed in Fig. 8 shows the width occupancy of the adulterated milk samples for each percentage of adulteration. It can be clearly seen that as the percentage of flour in milk is increased, the width occupied by the sample inside the channel is increased. It is a well-known fact that flour is denser than starch and urea and hence the viscosity value increases in larger proportions when compared to the other adulterants. The R2 value also suggests the linear regression adherence to the theoretical study of the two immiscible fluids flowing in the channel. The linear equation can be used to determine the value of the viscosity if the width occupied by the fluid in the channel is fed in.
176.04μm to 223.45μm for 0.5% adulteration to 10% adulteration respectively, thereby indicating the increase in the viscosity on the addition of starch to milk. A linear regression line was derived for measuring starch adulteration in various percentages. The range of is slightly lower than the range of the flour adulterated milk samples. This is a clear indicator that though starch is dense, but it is not as dense as the flour which can be calculated based on the width occupancy of 185.87 μm for the 0.5% adulteration of flour. The equation can be used to determine the viscosity of the adulterated milk by knowing the value of width occupied. There is also a difference in the visual inspection of the milk with flour and with starch. Starch, since it is pure white in color, goes unnoticed to a common man’s eyes, whereas the flour particles become vivid if they have not been added to hot water and mixed with milk [46]. There is also a notable point in the addition of starch to milk. Since starch contains a miniscule concentration of glucose, after the experimentation is performed one can find a lot of residue left over in the channel and proper flushing is to be done for the further use of the device [47]. D. Urea Adulteration in Milk In the case of urea adulteration in milk as shown in Fig. 10, it can be seen clearly that when 0.5% of urea is mixed with the milk then the width occupied by the sample in the channel was around 175 μm which is lower in comparison to flour (186 μm) and slightly lower in comparison to starch (176 μm). Based on the calculations a linear regression fitting line was derived for measuring adulteration of urea in milk for various percentages. This could be used to determine the value of the viscosity based on the width occupied by the sample. There is a close adherence of the values to the linear relation between the width occupied by the sample and the viscosity of the same. The closer the value of R2 is to 1, the higher the linear relation between them. As the content of urea in milk was increased from 0.5% to 5% there is sudden increase in occupancy of adulterated milk in Y shape micro channel device.
C. Starch Adulteration in Milk As shown in Fig. 9, the width occupied by the adulterated milk in the microviscometer slowly increases from
This type of analysis provides a distinctive method of viscosity calculation on a simple optofluidic microviscometer.
Fig. 9. Width occupied by starch adulterated milk sample vs sample viscosity and % of adulteration.
the water adulteration in milk in a more accurate and precise manner.
IV. C ONCLUSION
3006
IEEE SENSORS JOURNAL, VOL. 16, NO. 9, MAY 1, 2016
The experimental output clearly indicates that the real-time detection and monitoring of milk adulteration can be done using this device. According to the width value occupied by the individual fluids flowing in the device, the dynamic viscosity of the unknown sample can be calculated using the modified Hagen Poiseuille law. The concentration dependent viscosity of the adulterated milk can be found out based on occupancy of the sample in the microfluidic channel. The analysis proves that the addition of an adulterant in milk invariably affects its viscosity thereby making it easy to detect. Although the type of adulterant might be difficult to identify initially, but careful image analysis can easily show the presence of adulterants (in the case of flour or starch) in milk. In the case of water addition to the milk, a visual inspection of the thickness of milk will be a clear indicator of the adulteration. This device is accurate to measure even 1% of adulteration in milk thereby making it one of the most reliable and robust measurement device. This fully automated and robust device can be fabricated using the well-established 3D printing technique. It has high durability and is re-usable after flushing the reference fluid. Hence, the use of this device in other applications such as biodiesel blending and conventional fuel adulteration will be the next forward leap. The current device has constraints with respect to the type of adulterant and detection of two or more adulterants in milk. However, this is an attempt to realize a palm sized device for adulteration measurement in liquids. In the future scope, an electronic version of the microviscometer will be explored which will work on the similar principle i.e. for any fluid, time required to travel a unit distance in a micro-channel, of a given cross-section, is inversely proportional to its viscosity. There will be no reference fluid required for such a device and its electrical output can have higher possibilities of integration with other integrated control systems. This electronic version of the microviscometer can also be modified for other applications like food adulteration and hemoglobin detection in blood. ACKNOWLEDGMENT The authors would like to acknowledge the overall support and guidance of Dr. S J Chopra, Chancellor, UPES. R EFERENCES [1] A. S. Wiley, “Milk for ‘growth’: Global and local meanings of milk consumption in China, India, and the United States,” Food Foodways, vol. 19, no. 1, pp. 11–33, 2011. [2] A. Kumar, P. K. Joshi, P. Kumar, and S. Parappurathu, “Trends in the consumption of milk and milk products in India: Implications for self-sufficiency in milk production,” Food Security, vol. 6, no. 5, pp. 719–726, 2014. [3] Department of Animal Husbandry, Dairying & Fisheries, Ministry of Agriculture, Government of India, Statistics, BAHS-Basic Animal Husbandry. New Delhi, India: Krishibhavan, 2012. [4] S. Lingathurai and P. Vellathurai, “Bacteriological quality and safety of raw cow milk in Madurai, South India,” Microbiology, vol. 1, no. 10, 2010, Art. ID WMC001029. [5] J. Breman, Patronage and Exploitation: Changing Agrarian Relations in South Gujarat, India. Berkeley, CA, USA: Univ. of California Press, 1974. [6] S. Roy and D. V. Rangnekar, “Farmer participatory need-based extension (FPNE) approach: A sustainable model adopted by cooperative milk unions in Andhra Pradesh, India,” Livestock Res. Rural Develop., vol. 19, no. 10, 2007, Art. no. 144.
[7] G. Myrdal, “Corruption as a hindrance to modernization in South Asia,” in Political Corruption: A Handbook, A. J. Heidenheimer, M. Johnston, and V. T. Le Vine, Eds. New Brunswick, NJ, USA: Transaction Books, 1989, pp. 79–265. [8] L. Samal and A. K. Pattanaik, “Dairy production in India—Existing scenario and future prospects,” Int. J. Livestock Res., vol. 4, no. 2, pp. 105–113, 2014. [9] R. Chakravarty, “IT at milk collection centers in cooperative dairies: The national dairy development board experience,” in Information and Communication Technology in Development: Cases From India. Chennai, India: Sage Publications, 2000. [10] G. H. Richardson, “Automated testing of milk,” J. Dairy Sci., vol. 64, no. 6, pp. 1087–1095, 1981. [11] A. Sadat, P. Mustajab, and I. A. Khan, “Determining the adulteration of natural milk with synthetic milk using AC conductance measurement,” J. Food Eng., vol. 77, no. 3, pp. 472–477, Dec. 2006. [12] M. Sharifi and B. Young, “Towards an online milk concentration sensor using ERT: Correlation of conductivity, temperature and composition,” J. Food Eng., vol. 116, no. 1, pp. 86–96, 2013. [13] M. Kozempel, “Viscosity and density of lactulose solutions,” J. Dairy Sci., vol. 79, no. 12, pp. 2152–2154, 1996. [14] G. S. Bisht, S. Holmberg, L. Kulinsky, and M. Madou, “Diffusionfree mediator based miniature biofuel cell anode fabricated on a carbon-MEMS electrode,” Langmuir, vol. 28, no. 39, pp. 14055–14064, 2012. [15] T. H. Shellhammer, T. R. Rumsey, and J. M. Krochta, “Viscoelastic properties of edible lipids,” J. Food Eng., vol. 33, nos. 3–4, pp. 305–320, 1997. [16] J. Topolnicki, M. Kudasik, N. Skoczylas, and J. Sobczyk, “Low cost capillary flow meter,” Sens. Actuators A, Phys., vol. 152, no. 2, pp. 146–150, 2009. [17] B. B. Ghosh, S. K. Haldar, and A. Nag, “Synthesis of biodiesel from oils of jatropha, karanja and putranjiva to utilize in Ricardo engine and its performance & emission measurement,” in Proc. 4th BSME-ASME Int. Conf. Thermal Eng., Dec. 2008, pp. 731–738. [18] S. R. Yadav, V. K. Murthy, D. Mishra, and B. Baral, “Estimation of petrol and diesel adulteration with kerosene and assessment of usefulness of selected automobile fuel quality test parameters,” Int. J. Environ. Sci. Technol., vol. 1, no. 4, pp. 253–255, Dec. 2005. [19] D. Kashyap et al., “Application of electrochemical impedance spectroscopy in bio-fuel cell characterization: A review,” Int. J. Hydrogen Energy, vol. 39, no. 35, pp. 20159–20170, 2014. [20] M. Santhiago and L. T. Kubota, “A new approach for paper-based analytical devices with electrochemical detection based on graphite pencil electrodes,” Sens. Actuators B, Chem., vol. 177, pp. 224–230, Feb. 2013. [21] S. Xiong, A. Q. Liu, L. K. Chin, and Y. Yang, “An optofluidic prism tuned by two laminar flows,” Lab Chip, vol. 11, no. 11, pp. 1864–1869, 2011. [22] T. Kobayashi and S. Konishi, “Microfluidic chip with serially connected filters for improvement of collection efficiency in blood plasma separation,” Sens. Actuators B, Chem., vol. 161, no. 1, pp. 1176–1183, 2012. [23] M. H. Sun et al., “Characterization of microfluidic fuel cell based on multiple laminar flow,” Microelectron. Eng., vol. 84, nos. 5–8, pp. 1182–1185, 2007. [24] J. W. Weaver, S. A. Skaggs, D. L. Spidle, and G. C. Stone, “Composition and behavior of fuel ethanol,” US Environ. Protection Agency, Office Res. Develop., Washington, DC, USA, Tech. Rep. EPA/600/R-09/037, 2009. [25] N. K. L. Wiziack et al., “A sensor array based on mass and capacitance transducers for the detection of adulterated gasolines,” Sens. Actuators B, Chem., vol. 140, no. 2, pp. 508–513, 2009. [26] Z. Han, X. Tang, and B. Zheng, “A PDMS viscometer for microliter Newtonian fluid,” J. Micromech. Microeng., vol. 17, no. 9, p. 1828, 2007. [27] Z. Han and B. Zheng, “A poly(dimethylsiloxane) viscometer for microliter power law fluids,” J. Micromech. Microeng., vol. 19, no. 11, p. 115005, 2009. [28] J. Chevalier and F. Ayela, “Microfluidic on chip viscometers,” Rev. Sci. Instrum., vol. 79, no. 7, p. 076102, 2008. [29] S. Goel, P. S. Venkateswaran, R. Prajesh, and A. Agarwal, “Rapid and automated measurement of biofuel blending using a microfluidic viscometer,” Fuel, vol. 139, pp. 213–219, Jan. 2015. [30] M. Gorji, M. Alipanah, M. Shateri, and E. Farnad, “Analytical solution for laminar flow through leaky tube,” Appl. Math. Mech., vol. 32, no. 1, pp. 69–74, 2011.
VENKATESWARAN et al.: RAPID AND AUTOMATED MEASUREMENT OF MILK ADULTERATION
[31] L. M. Reid, C. P. O’Donnell, and G. Downey, “Recent technological advances for the determination of food authenticity,” Trends Food Sci. Technol., vol. 17, no. 7, pp. 344–353, 2006. [32] M. A. Martínez, “Hagen-Poiseuille flow solutions in Grad-type equations,” J. Statist. Phys., vol. 142, no. 4, pp. 710–725, 2011. [33] K. Dutkowski, “Experimental investigations of Poiseuille number laminar flow of water and air in minichannels,” Int. J. Heat Mass Transf., vol. 51, nos. 25–26, pp. 5983–5990, 2008. [34] O. H. Paydar, C. N. Paredes, Y. Hwang, J. Paz, N. B. Shah, and R. N. Candler, “Characterization of 3D-printed microfluidic chip interconnects with integrated O-rings,” Sens. Actuators A, Phys., vol. 205, pp. 199–203, Jan. 2014. [35] P. Di Sia, “Present and future of nanotechnologies: Peculiarities, phenomenology, theoretical modelling, perspectives,” Rev. Theoretical Sci., vol. 2, no. 2, pp. 146–180, 2014. [36] J. Xuan, M. K. H. Leung, D. Y. C. Leung, and H. Wang, “Laminar flow-based fuel cell working under critical conditions: The effect of parasitic current,” Appl. Energy, vol. 90, no. 1, pp. 87–93, 2012. [37] R. B. Bird, W. E. Stewart, and E. N. Lightfoot, Transport Phenomena, 2nd ed. New York, NY, USA: Wiley, 2002. [38] B. Bhushan and O. Marti, “Scanning probe microscopy–principle of operation, instrumentation, and probes,” in Springer Handbook of Nanotechnology. Berlin, Germany: Springer, 2010, pp. 573–617. [39] M. A. Ibrahim and H. Elhaes, “Exploring materials: Molecular modeling approach,” Rev. Theoretical Sci., vol. 1, no. 3, pp. 368–376, 2013. [40] L. Wang, R. Kodzius, X. Yi, S. Li, Y. S. Hui, and W. Wen, “Prototyping chips in minutes: Direct laser plotting (DLP) of functional microfluidic structures,” Sens. Actuators B, Chem., vol. 168, pp. 214–222, Jun. 2012. [41] C.-C. Chang and R.-J. Yang, “Electrokinetic mixing in microfluidic systems,” Microfluidics Nanofluidics, vol. 3, no. 5, pp. 501–525, 2007. [42] T. F. Holden, N. C. Aceto, and E. F. Schoppet, “Effects of viscosity and temperature on the foaming characteristics of concentrated whole milk,” J. Dairy Sci., vol. 47, no. 4, pp. 359–364, 1964. [43] S. Kasemsumran, W. Thanapase, and A. Kiatsoonthon, “Feasibility of near-infrared spectroscopy to detect and to quantify adulterants in cow milk,” Anal. Sci., vol. 23, no. 7, p. 907, 2007. [44] K. C. Sahu, H. Ding, P. Valluri, and O. K. Matar, “Linear stability analysis and numerical simulation of miscible two-layer channel flow,” Phys. Fluids, vol. 21, no. 4, p. 042104, 2009. [45] M. Ayub, Q. Ahmad, M. Abbas, I. M. Qazi, and I. A. Khattak, “Composition and adulteration analysis of milk samples,” Sarhad J. Agricult., vol. 23, no. 4, pp. 1127–1130, 2007. [46] S. K. Nayak et al., “Heat stability and flow behaviour of buffalo milk added with corn starch,” Food Hydrocolloids, vol. 18, no. 3, pp. 379–386, 2004. [47] N. Silalai and Y. H. Roos, “Mechanical relaxation times as indicators of stickiness in skim milk–maltodextrin solids systems,” J. Food Eng., vol. 106, no. 4, pp. 306–317, 2011.
Pedinti Sankaran Venkateswaran received the B.E. degree in electronics and communications from Anna University, Chennai, India. He received the Gold Medal for the M.Tech. degree in nanotechnology from Anna University, Chennai, in 2011. He has submitted his thesis for the award of Doctor of Philosophy under the joint supervision of Dr. Santosh Dubey (UPES) and Dr. Sanket Goel (BITS) at the University of Petroleum and Energy Studies, Dehradun, India, majoring in microfluidic lab-on-chip devices. He is currently working at ARK Infosolutions pvt Ltd., Bangalore, India, in the capacity of Manager (Research & Development) and Product Manager (3D Printing Solutions). He has close to five years of experience in research management and administration with a specific emphasis on strategy and planning. His research interests concentrate on micro and nanotechnology with a focus on interdisciplinary fields, such as 3D modeling, 3D printing, rapid prototyping, microfluidics, lab-on-chip devices, and sustainable energy.
3007
Abhishek Sharma received the bachelor’s degree in electronics and communication engineering from ITM-Gwalior, India, in 2012, and the master’s degree in robotics engineering from the University of Petroleum and Energy Studies (UPES), Dehradun, India, in 2014. He was a Senior Research Fellow in a DST funded project under the Technology Systems Development Scheme. He is an Assistant Professor with the Department of Electronics and Instrumentation, UPES. His research interests include embedded system, and control and robotics. Santosh Dubey received the M.S. and Ph.D. degrees from the Department of Scientific Computing, Florida State University, Tallahassee, USA, in 2008 and 2012, respectively. He held a post-doctoral position with the School of Nuclear Engineering, Purdue University, West Lafayette, IN, USA. Before going to the U.S. for higher studies, he was a Scientist with the Institute for Plasma Research, Department of Atomic Energy, India, from 2001 to 2006. He has been an Assistant Professor with the Department of Physics, University of Petroleum and Energy Studies, Dehradun, India, since 2013. He employs modeling and simulation schemes to different areas of interest including but not limited to material science, nuclear engineering, reservoir modeling, photovoltaics, and microfluidics. Ajay Agarwal (SM’15) received the bachelor’s degree from the National Institute of Technology, Rourkela, India, and the M.S. and Ph.D. degrees from the Birla Institute of Technology and Science, Pilani, India. Earlier as a member of the Technical Staff, he served the Institute of Microelectronics, Singapore, for over nine years. His engagement with semiconductor industries and research institutes is for 24 years. He is currently a Senior Principal Scientist and Nodal Officer of the Central Electronics Engineering Research Institute, Pilani. He is involved in the development of nanotechnologies MEMS, and microsensors. He is also an Associate Professor with the Academy of Scientific and Innovative Research, New Delhi, India. He has authored over 210 research publications, 25 invited/plenary/keynote talks, and 25 patents. He is a Lifetime Fellow of MSI (India). His major research interests are the development of micro/nanotechnologies, MEMS, and semiconductor processes for various applications. He is bestowed with various awards, including the 2008 National Technology Award-Singapore, the 2009 Excellence Award, the IME Singapore; Collaboration Development Award from the British High Commission (Singapore) in 2005 and 2006, and Super Kaizen (four times) and Best Kaizen (seven times) at USHA (India) Ltd.
Sanket Goel received the B.Sc. (H-physics) degree from Ramjas College, Delhi University, the M.Sc. (physics) degree from IIT Delhi, and the Ph.D. (electrical engineering) degree from the University of Alberta, Canada, under the NSERC Fellowship. He headed the R&D Department and worked as an Associate Professor with the Electronics and Instrumentation Engineering Department, University of Petroleum and Energy Studies, Dehradun, India, from 2011 to 2015. He is currently an Associate Professor with the Birla Institute of Technology and Science–Pilani. He has close to 40 publications, four patents (one U.S. and three Indian) to his credits, delivered several invited talks, and guided three Ph.D. and nine master’s students. His current research interests are microfluidics and nanotechnology, materials and devices for energy (both conventional and renewable), and biomedical applications. He has won several awards during the course of his career, including the Prestigious Fulbright-Nehru Fellowship (2015), the Young Scientist Award (2013), the Best Students Paper Award (2005), and the Ph.D. Thesis Award (2005).
