Computerized, Transient Hot-Wire Thermal

Conductivity (HWTC) Apparatus

For Nanofluids

The 6th WSEAS International Conference on HEAT and MASS TRANSFER

(WSEAS - HMT'09)
Ningbo, China, January 10-12, 2009

M. Kostic & Kalyan C. Simham

Department of Mechanical Engineering
NORTHERN ILLINOIS UNIVERSITY

www.kostic.niu.edu
 Overview
 INTRODUCTION
 OBJECTIVE
 THEORY OF HOT-WIRE METHOD
 PRACTICAL APPLICATION OF HOT-WIRE
METHOD
 DESIGN OF HOT-WIRE CELL
 INSTRUMENTATION
 DATA ACQUISTION
 CALIBRATION
 UNCERTAINTY ANALYSIS
 RESULTS
 CONCULSIONS
 RECOMMENDATIONSwww.kostic.niu.edu
 INTRODUCTION
 Nanofluids are colloidal suspensions of
nanoparticles, nanofibers, nanocomposites in
common fluids
 They are found to have enhanced thermal
properties, especially thermal conductivity
 Thermal conductivity values of nanofluids may be
substantially higher than related prediction by
classical theories
 No-well established data or prediction formula
suitable to all nanofluids
 Experimental thermal conductivity measurement of
nanofluids is critical
www.kostic.niu.edu
Table 1: Summary of landmark development in nanofluids
*

www.kostic.niu.edu
* (reprinted with permission; reference listed within this table are with respect to (Manna et al 2005))
 Nanofluid Preparation Methods

• One Step (Direct

Evaporation and
Condensation) Method
• Two Step Method or Kool-
aid Method

• Chemical Method Fig1: Improved new-design for the one-step, direct

evaporation-condensation nanofluid production
apparatus, (Kostic 2006)

www.kostic.niu.edu
Thermal Conductivity
• Material Property
• Determines ability to conduct heat
• Important for thermal Management

Classification of Thermal Conductivity

Measurement Techniques for Fluids
Horizontal Flat Plate Method
Steady State Methods Vertical Coaxial Cylinder Method
Steady State Hot-Wire Method
Line Source (Hot-Wire)
Method Source Method
Cylindrical
Non-Steady State
Spherical Source Method
Methods www.kostic.niu.edu
Plane Source Method
Transient Hot-Wire Method for Fluids
Advantages:
• Fast and Accurate
• Minimum Conduction and Radiation losses
• Minimize (or even avoid) Convection

Classification of Hot-Wire Methods

• Standard Cross Wire Method
• Single Wire, Resistance Method
• Potential Lead Wire Method
• Parallel Wire Method
www.kostic.niu.edu
 OBJECTIVE
 Device to Suspend Hot-Wire
 Reduce Nanofluid Sample Size
 Minimize End Errors
Design  Uniform Tension on Hot-Wire
 Separate Wires for Power and
Voltage
 Monitor Temperature
 Mechanism to Calibrate Hotwire
Tension
 Flexibility for Cleaning and Handling
www.kostic.niu.edu
 OBJECTIVE

 Electrical Circuit
Instrumentation
 Flexible Connections

 Optimize to Reduce
Data Acquisition Noise
 Develop Program
Calibration  Standard Fluids
Uncertainty
 Thermal Conductivity
Analysis www.kostic.niu.edu
Principle of Hot-Wire Method
• An infinitely long and thin, ideal continuous line source
dissipating heat into an infinite medium, with constant heat
generation
General Fourier’s Equation
1 T 1   T 
 r 
 f t r r  r 
Where
Boundary Conditions T  T  T0
  T  q T is the final temperature,
t  0 and r  0 limr     
r 0
  r  2k f T0 is the initial temperature,
r is the radial distance and
t0 and r   lim T  r , t   0 t is the time
r 
q is heat flux
 f is thermal diffusivity
Ideal case: kf is Thermal Conductivity
Line source has an infinite thermal conductivity
and zero heat capacity www.kostic.niu.edu
• The temperature change at a radial distance r, from the heat source is
conforms to a simple formula by applying boundary conditions
q  r2 
T  r , t   T  r , t   T0  Ei 
4k f  4 f t 


series expansion of the exponential integration

  r 2   r 2 2 
     
q   4 f t    4 f t   4 f t   
T  T (r , t )  T0     ln 2      ..  ....... 
4k f   r   1  1! 2  2! 
  
   

Where,  =0.5772 is the Euler’s constant

• At any fixed radial distance, in two instances in time the equation,
the temperature change can be represented as
q t 
T2  T1  ln 2 
4k f  t1 
www.kostic.niu.edu
 • A plot of temperature against the natural logarithm of time results
in a straight line, the slope being propositional to kf
Thermal Conductivity
q d ln(t )
kf 
4 dT

Practical application of hot-wire method

• The ideal case of continuous line is approximated with a
finite wire embedded in a finite medium

www.kostic.niu.edu
Figure 2.1 Typical plot of temperature change against time for hot-wire experiment (Johns et al 1988)
 Nanofluids Thermal Conductivity
Methods By Other Authors
Nanofluid Thermal Conductivity
Author, Year
Measurement Method
Wang et al (1999) Horizontal flat plate method

Lee et al (1999), Yu et al
(2003) and Vadasz Vertical, single wire, hot-wire method
(2006)

Assael et al (2004) Two wires, hot-wire method

Manna et al (2005) Thermal comparator

Ma (2006) Horizontal, single wire, hot-wire method

Simham (2008) Vertical, single wire, hot-wire method

www.kostic.niu.edu
Hot-wire Method for Nanofluid
• Nanofluids are electrically conducting fluids
• Availability of nanofluids
• Thermal expansion of wire
• Cleaning of the cell

Hot-Wire Method for Electrically Conducting Fluids

Problems identified by Nagasaka and Nagashima (1981)
• Possible current flow through the liquid, resulting in
ambiguous measurement of heat generated in the wire,
• Polarization of the wire surface,
• Distortion of small voltage signal due to combination of
electrical system with metallic cell through the liquid.
www.kostic.niu.edu
 q  1 
T 
4k f ln t  Ao  t  Bo ln t  C o  
 
Where,
 4 f  2k f ro kf
Ao  ln 2 
 k ln   ...
 ro   i rw 2k w

1  2  k i k w  2  k f k 
Bo  rw     ro   
i

   i  w   
2k f   f  i 

rw2  k f  k i  1 1  4 2  ro2  1 1  rw2  k i k w   ro 

Co               ln 
 
8  k w   w  i   i  w  2  f  i  ki   i  w   rw 

1  2  k i k w  2  k f k i  4 f
 rw     ro    ln 2
     
2k f   i w   f i  ro 

1 t  Bo ln t  C o  is due to the presence of the insulation layer on the wire

Ao shifts (i.e. offsets) the plot of Tagainst ln (t), without changing the slope
www.kostic.niu.edu
 Insulation Coating Influence on Thermal
Conductivity Measurement
Yu and Choi
(2006)
• The results of numerical simulation and experimental test
show that, for most of the engineering applications, the
relative measurement error of the thermal conductivity
caused by the insulation coating are very small if the slopes
of the temperature rise – logarithmic time diagram are
calculated for large time values
• No correction to insulation coating is necessary even for the
conditions that the insulation coating thickness is comparable
to the wire radius, and that the thermal conductivity of the
insulation coating is lower than that of the measured medium
www.kostic.niu.edu
 Reasons For Adapting Single Wire Method

• Simplicity of Operation
• Low Cost
• Easy Insulation Coating
• Easy Construction
• Design Optimized

www.kostic.niu.edu
 Design Parameters
• Size of the wire (i.e., Wire radius)
• Type of insulation coating
• Length of the wire
• Sample size (length and radius of the cell)
Selected Design Parameters
• Wire Diameter 50.8 µm
• Teflon Insulation coating thickness 25.4 µm
• Measured length of wire (after fabrication) is 0.1484
m
• Diameter of bounding wall is 0.0144 m
• Length of sample is 0.165 m
www.kostic.niu.edu
www.kostic.niu.edu
 Spring Rod with Threading Calibration Gauge
(to guard spring rod and
Locking Nut
calibrate the spring tension)
(calibrated weight for required
spring tension) To the Data Acquisition System
Power Supply Connector
Connectors and
Calibration Guage Holder

Special Shape Sliding Fit Hole D-Type Connector

(avoids turning of spring)
Cell Cap with Rectangular Cuts Hot-Wire Voltage Output Wires
(for wire outlet)
T-Type Thermocouples
Tension Spring
(spring constant 0.02 N/mm) Wire Holder

Constant Voltage Input Wires

Striped Stranded Copper Wire
(to provide flexiblity and avoid backlash)
Sliding Tube
(aligns the hot-wire)
Hot-Wire Guiding Block
(off-centered)

Inner Wire Guide

Wire Protection Clip # 1

Soldered Joint # 1

Outer Shell
(test-fluid reservoir)

Teflon Coated Platinum Hot-Wire

Ø 0.0508 mm Inner Semi-Circular
Coating Thickness 0.0245 mm Hot-Wire Holder
Measurement Section
149.2 mm

Wire Protection Clip # 2

Threaded Nut

Soldered Joint # 2 Thermocouple at the Bottom

L45°
Off-Centered Alignment Ring Wire Protection Clip # 3

Teflon Sealing Insulated Copper Wire

Ø 0.254 mm
Threaded Hole in Base Plate
Cell Base Plate (Assembly and Cleaning)

www.kostic.niu.edu
Fig 2: Cross-sectional front view of improved transient hot-wire thermal
 Spring Rod with Threading Calibration Gauge
(to guard spring rod and
Locking Nut
calibrate the spring tension)
(calibrated weight for required
spring tension) To the Data Acquisition System
Power Supply Connector
Connectors and
Calibration Guage Holder

Special Shape Sliding Fit Hole

D-Type Connector
(avoids turning of spring)
Cell Cap with Rectangular Cuts Hot-Wire Voltage Output Wires
(for wire outlet)
T-Type Thermocouples
Tension Spring
(spring constant 0.02 N/mm) Wire Holder

Constant Voltage Input Wires

Striped Stranded Copper Wire
(to provide flexiblity and avoid backlash)
Sliding Tube
(aligns the hot-wire)
Hot-Wire Guiding Block
(off-centered)

Inner Wire Guide

Fig 2: Top half cross-sectional front view of transient hot-wire thermal conductivity cell
Wire Protection Clip # 1
Soldered Joint # 1 www.kostic.niu.edu
 (aligns the hot-wire)
Hot-Wire Guiding Block
(off-centered)

Inner Wire Guide

Wire Protection Clip # 1

Soldered Joint # 1

Outer Shell
(test-fluid reservoir)

Teflon Coated Platinum Hot-Wire

Ø 0.0508 mm Inner Semi-Circular
Coating Thickness 0.0245 mm Hot-Wire Holder
Measurement Section
149.2 mm

Wire Protection Clip # 2

Threaded Nut

Soldered Joint # 2 Thermocouple at the Bottom

L45°
Off-Centered Alignment Ring Wire Protection Clip # 3

Teflon Sealing Insulated Copper Wire

Ø 0.254 mm
Threaded Hole in Base Plate
Cell Base Plate (Assembly and Cleaning)

www.kostic.niu.edu
Fig 3: Bottom half cross-sectional front view of transient hot-wire thermal conductivity cell
 Thermocouple
Off-Centered
at the middle
Alignment Ring

Semi-Circular Base Plate

Hot-Wire Holder
(Off Centered)
Threaded Nut

Outer Shell
Ø14.371mm
(test-fluid reservoir)

Ø17.424mm
Protection Clip

Fig 4: Cross sectional top view of the hot-wire cell at the middle

www.kostic.niu.edu
 Sliding Hole
Calibration Guage
D-Type Connector
Locking Nut (thermocouples and wire voltage
(calibrated weight fpr required measurement using
spring tension) data acquisition system)

Power Supply Connector Connectors and

Calibration Guage Holder
Fixing Nut

Tension Spring
T-Type Thermocouple
Hot-Wire Voltage Output Wires
Constant Voltage Input

Wire Holder
Wire Holder Fixing Nut

Outer Shell
(test-fluid reservoir)

Threaded Nut
(soldered to outer shell )
Cell Base Plate

www.kostic.niu.edu
Fig 5: Isometric view of transient hot-wire thermal conductivity cell
 Tension Spring
Locking Screw
(avoids the axial movement
of calibration guage)

Sliding tube
Thermocouple at the Top
L15°
Rectangular hole
on the Inner Cell
(for guiding the wires out)

Thermocouple at the Middle

L75°

Wire Guiding Hole

(to guide the aligned wires out) Thermocouple at the Bottom
L45°
Off-Centered Alignment Ring
www.kostic.niu.edu
Fig 6: Left-side view of transient hot-wire thermal conductivity cell without the outer cell, base
plate and protection pins
 Off-Centered
Alignment Ring
(Provides Rigidity
to the other end
of the wire)

Uniform Tension
on the Platinum Wire

Sliding Tube
(causes free movement
without friction)

Tension Spring
(spring constant
0.02 N/mm)

Calibration Guage Calibrated Weight

(guards the spring rod (for required spring tension
and protects the platinum wires within elastic modulus
for sudden shocks) of the platinum hot-wire)

www.kostic.niu.edu
Fig 7: Calibration position of the hot-wire cell
 (1) Spring Rod
ΔZcal

Spring Assembly
Zcal

ΔZ0 (2) Locking Nut

Cell Cap

Fwa  W1  W2  W3  W4
Where,
Weight of spring rod,W1 = 0.00708 N
(3) Tension Spring

Spring Constant ζs Weight of locking nut, W2 = 0.1762 N

Initial Spring Force Fsi
Weight of tension spring,W3 = 0.0115 N
Weight of sliding tube,W4 = 0.00490 N

(4) Sliding Tube Fwa = 0.1997 N

W1  W2  W3   Fsi
Z cal 
s
Fwa
Z cal = 0.0056 m
www.kostic.niu.edu
 www.kostic.niu.edu
Fig 8: Fabricated transient hot-wire thermal conductivity apparatus cell
 Instrumentation

www.kostic.niu.edu
Figure 5.1 Schematics of electrical circuit with data acquisition system
Measurement Procedure
• The wire is heated with electrical constant power supply at
step time
• The wire simultaneously serves as the heating element and as
the temperature sensor
• The change in resistance of the wire due to heating is
measured in time using a Wheatstone bridge circuit
• The temperature increase of the wire is determined from its
change in resistance
• Thermal conductivity is determined from the heating power
and the slope of temperature change in logarithmic time

www.kostic.niu.edu
 Signal Analysis
Bridge Balance R1 Rw0

R2 R3

Resistance of the hot wire

R1
Rw0  R3
R2
The bridge voltage output
 Rw0  Rw R1 
Vout  Vin   
R
 w0  R w  R3 R1  R 2 

The Resistance change of Hot-Wire

  Vout 
R3  R1   R1  R2   
  Vin 
R wt 
  Vout 
 R2   R1  R2   
  Vin 
www.kostic.niu.edu
The Temperature change of Hot-Wire
1 Rwt 2  Rwt1 Rw
T  
 TCR Rw 0  TCR Rw0

The Voltage Drop Across the Hot-Wire

Vin R wt
V Rw 
R wt  R3

Heat Flux per Unit Length at any Instant of Time

VRw  2
q
Lw Rwt

Thermal Conductivity
q d ln(t )
kf 
4 dT

www.kostic.niu.edu
Computerized Data Acquisition

• Data acquisition hardware and software are optimized to minimize

signal noise and enhance gathering and processing of useful data

Types of Data Measured

• Bridge voltage output
• Bridge voltage input
• Hot-wire Voltage
• Temperature of fluid

Programming in LabVIEW
• A program has been written in LabVIEW application software to
automatically calculate thermal conductivity

www.kostic.niu.edu
 Data Acquisition Hardware
• PCI – 6024E, Multifunctional DAQ Board
(E–series family, PCI, PCMCIA bus, 16 single-ended/ 8 differential channel
analog inputs, 12 bit input resolution, 200 kS/s maximum sampling rate, ±
0.05 V to ± 10 V input range, 2 analog inputs, 12 bit output resolution, 10
kSamples/s output range, 8 digital I/O, two 24 bit counter timer, digital
trigger)
• SCXI – 1000, 4 Slot Signal Conditioning Chassis
(shielded enclosure for SCXI module, low – noise environment for signal
conditioning, forced air cooling, timing circuit)
• SCXI – 1102, 32 Differential Channel Thermocouple Input Module
(programmatic input range of ± 100 mV to ± 10 V per channel, overall gain
of 1 – 100, hardware scanning of cold junction sensor, 2 Hz low pass filtering
per channel, relay multiplexer, over voltage protection of ± 42 V, 333 kS/s
maximum sampling rate, 0-50 ºC operation environment temperature)
• SCXI – 1303, 32 Channel Isothermal Terminal Block for
Thermocouple modules
(SCXI front end mountable terminal block for SCXI-1100 and SCXI-
1102/B/C, cold junction compensation sensor, open-thermocouple detection
circuitry, isothermal construction for minimizing errors due to thermal
gradient, cold junction accuracy for 15-35 ºC is 0.5 ºC and for 0-15 ºC & 25-
www.kostic.niu.edu
50 ºC is 0.85 ºC, repeatability is 0.35 ºC)
 Data Acquisition Hardware
• SCXI – 1122, 16 Differential Channel Isolated Universal Input
Module
(DC input coupling, nominal range ± 250 V to ± 5 mV with overall gain of
0.01 to 2000, over voltage protection at 250 Vrms, maximum working voltage
in each input should remain with 480 Vrms of ground and 250 Vrms of any
other channel, cold junction compensation, bridge compensation, isolated
voltage and current excitation, low pass filter setting at 4 kHz or 4 Hz, shunt
calibration, 16 relay multiplexer, 100 Samples/s (at 4 kHz filter) and 1
Sample/s (at 4 Hz filter), two 3.333 V excitation level sources)
• SCXI – 1322, Shielded Temperature Sensor Terminal Block
(SCXI front end mountable terminal block for SCXI -1122, on board cold
junction sensor)
• SCXI – 1349, Shielded Cable Assembly
(adapter to connect SCXI systems to plug-in data acquisition devices, mounting
bracket for secure connection to the SCXI chassis)
• SH68-68-EP, Noise Rejecting, Shielded Cable
(Connects 68-pin E Series devices (not DAQ cards) to 68-pin accessories,
individually shielded analog twisted pairs for reduced crosstalk with high-speed
boards)
www.kostic.niu.edu
 Start

Bridge
Balance

Measure
Resistance

Calculate Initialize Variables

Initial Resistance R w0 R1, R 2, R ,3 TCR, L w

Measurement Time
Measure Temperature at
Average Temperature
Top, Middle and Bottom
of the Cell
of Hot-Wire Cell

Voltage Output V out

Sampling Rate

If
NO YES
Vout >= 0.001 V

Measure Bridge
Plot V outVs Ln(t) Output Voltage V out
and time t

Measure Bridge
Input Voltage V in

Measure Temperature at
Average Temperature
Top, Middle and Bottom
of the Cell
of Hot-Wire Cell

Calculate Total Calculate

Change in Resistance Wire Voltage V Rw

Plot Calculate
Temperature Vs Ln(time) Change in Temperature

Calculate Slope of Calculate

Time Ranges Temperature and Ln(Time) Heat Input
for Specified Time Range per unit length 'q'

Calculate
Thermal Conductivity

Store
All
Data

www.kostic.niu.edu
End

Figure 5.3: LabVIEW Program Algorithm for Thermal Conductivity Measurement

 Calibration
Two Standard Fluids Ethylene Glycol and Water

Reference Temperature
1
Tr  T0   T  t1   T  t 2  
2

, Resistances of the Wheatstone bridge circuit are measured as

R1 = 2270.6 Ω Lw = 0.1484 m
R2 = 2161.1 Ω  TCR  Z  R w
R3 = 7.715 Ω Where,
Rw0 = 8.106 Ω Z  = 0.02652 Ω/°C
Z  is the the slope of dRw vs T
www.kostic.niu.edu Rw = 8.22 Ω
 16

Ethylene Glycol 14
Distilled Water
Log. (EG (2.0s - 6.0s))
Log. (Water (2.0s-6.0s)) 12
Wire Temperature Change, ΔT [°C]

10

0
0.01 0.1 1 10 100
time, t [s]

Figure 6.1: Wire temperature www.kostic.niu.edu

change against time (in logarithmic scale)
for ethylene glycol and distilled water
 Heat Input per Unit Length in Time
5.57

Water
Ethylene glycol
5.565
Heat Input per Unit Length, q [W/m]

5.56

5.555

5.55

5.545

5.54
0 5 10 15 20 25 30 35 40 45 50
Time, t [s]

www.kostic.niu.edu
Figure 6.2: Heat input per unit length against time (for ethylene glycol and water)
 Calibration Data from (1 s-10 s)

14

Ethylene Glycol
13 Distilled Water

12
Wire Temperature Change, ΔT [°C]

11

10

9
Valid time range for data reduction

5
1 10
time, t [s]

Figure 6.3: Calibration data from time (1 s – 10 s), shows the

www.kostic.niu.edu
selected time range for data reduction as 2s – 6 s, for ethylene glycol and water
 Repeatablity of Ethylene Glycol Thermal Conductivity Measurement

0.265 Repeatability of EG
Linear (Reference Value)
Linear (Mean)
Ethylene Glycol Thermal Conductivity, kfeg [W/m°C]

0.260

0.255

0.250

0.245

0.240
0 1 2 3 4 5 6 7 8 9 10
Measurement Set

Figure 6.4: Results of repeatability measurement of thermal conductivity for

www.kostic.niu.edu
Ethylene glycol, shows the bias and precision error in measurement
 Repeatability of Water Thermal Conductivity Measurement

0.650 Repeatablity of Water

Linear (Reference)
Linear (Mean)
0.640
Water Thermal Conductivity, kfw [W/m°C]

0.630

0.620

0.610

0.600

0.590

0.580
0 1 2 3 4 5 6 7 8 9 10
Measurement Set

Figure 6.5: Results of repeatability measurement of thermal conductivity

www.kostic.niu.edu
for distilled water, shows the bias and precision error in measurement
 Calibration Results

Table 6.1: Uncertainty in repeatability of measured thermal conductivity

Precision
Reference Measured
Fluid Bias Error Error Uncertainty
[W/m°C] [W/m°C]
(95 %)

Ethylene
Glycol 0.254 0.253 - 0.395 % 2.03 % 2.06 %
(32.5 °C)
Distilled
water 0.612 0.619 1.2 % 2.23 % 2.52 %
(~ 26 °C)

www.kostic.niu.edu
 Uncertainty in Thermal
Conductivity
q d ln(t )
kf 
4 dT

Rearranging in terms of the measured resistance change in the wire

q d ln(t ) q 1
k f  TCR Rw0  TCR Rw0
4 d Rw 4 Z R

Uncertainty

2 2 2 2
 k f   k f   k f   k f 
uk f   u q    u TCR    u Rw 0    uZR 
 q    TCR   Rw0   Z R 

www.kostic.niu.edu
Uncertainty in Heat Input per Unit Length
2 2 2
 q   q   q 

uq   uVRw    u Rwt    u Lw   Pq2
 VRw   Rwt   Lw 

Pq is the precision error in the average heat input per unit length

u q q  1.63 %.

Uncertainty in Wire Voltage

2 2 2
 V   V   VRw 
uVRw   Rw uVin    Rw u Rwt    u R3 
 Vin   Rwt   R3 
uVRw V Rw  0.706 %.
www.kostic.niu.edu
 Uncertainty in Total Resistance Change
2 2 2 2 2
 Rwt   Rwt   Rwt   Rwt   Rwt 
u Rwt   u R1    u R2    u R3    uVin    uVout 
 R1   R2   R3   Vout   Vin 

u Rwt Rwt  0.813 %.

Uncertainty in Measured Bridge Voltage Input

uVin Vin  0.535 %

Uncertainty in Measured Bridge Voltage Output

uVout Vout  1.0 %

www.kostic.niu.edu
 Uncertainty in Resistances
Uncertainty in Multimeter  u d  mm   u 0  2mm   u c  2mm
Uncertainty in Resistance R1 u R1   u d  2mm  BR2
1
u R1 R1  0.1 %.

Uncertainty in Resistance R2 u R2   u d  2mm  BR2 2

u R2 R2  0.25 %.

Uncertainty in Resistance R3 u R3   u d  2mm  BR2 3

u R3 R3  0.516 %.

Uncertainty in Resistance R3
2 2 2
 R   R   R 
u Rw 0   w u R1    w u R2    w u R3   B Rw 0
 R1   R2   R3 

u Rw 0 Rw0  1.63 %.
www.kostic.niu.edu
Uncertainty in Temperature Coefficient of Resistance
2 2
     
uTCR   TCR u Z     TCR u Rw  u TCR  TCR  2.275 %.
 Z    Rw 

Hot-Wire Resistance Vs Temperature

8.95

8.9

8.85
Hot-Wire Resistance Rw [Ω]

8.8

8.75

8.7

8.65
Rw = 0.026521 T + 7.698728
8.6 r2 = 0.999036

8.55

8.5

8.45
29 31 33 35 37 39 41 43 45

www.kostic.niu.edu
Temperature, T [°C]

Figure 6.7 Calibration of Temperature Coefficient of Resistance of Teflon Coated Platinum Hot-Wire
 Uncertainty in Length of Hot-Wire

u Lw   u d  VC
2
  Le    LFS 
2 2
u Lw Lw  0.0661 %.

Uncertainty in Slope of Total Resistance Change

against Logarithmic Time

  S yx  Z
NR
 S a1  Z 2
 NR 
R R NR
N R   ln t      ln t  i 
2
i
i 0  i 0 

u Z R  t 200,95%  S a1  Z R

u Z R Z R  0.2314%

www.kostic.niu.edu
Table 7.2: Percentage
uncertainties
Uncertainty (%)

uq q 1.629

u TCR  TCR 2.274

u Rw 0 Rw0 1.627

uZR Z R 0.231

uk f k f 3.245

www.kostic.niu.edu
 Nanofluid thermal conductivity
Measurement
Nanoparticles:
• Copper, particle size 35 nm
Base Fluid:
• Ethylene glycol and
Concentration: Water
• 1 volumetric %
Physical Stabilization:
• Ultrasonication
www.kostic.niu.edu
 Copper in Ethylene Glycol Nanofluid
Measured Thermal Conductivity Ratio of
1 vol% of Copper in Ethylene Glycol Nanofluid

1.16

1.14
Thermal Conductivity ratio knfeg/kfeg

1.12

1.1
1% vol Cu in EG
1.08 Linear (Mean)

Mean= 1.1282
1.06

1.04

1.02

1
0 1 2 3 4 5 6
Measurement Set
www.kostic.niu.edu
Figure 7.1: Nanofluid thermal conductivity measurement of 1 vol % of copper in ethylene glycol
 Copper In Water Nanofluid
Measured Thermal Conductivity Ratio of
1 vol% of Copper in Water Nanofluid
1.3500 1% vol Cu in Water
Linear (Mean)

1.3000 Mean = 1.1595

Thermal conductivity Ratio knfw/kfw

1.2500

1.2000

1.1500

1.1000

1.0500

1.0000
0 1 2 3 4 5 6
Measurement Set
www.kostic.niu.edu
Figure 7.2: Nanofluid thermal conductivity measurement of 1 vol % of copper in water
 Improvements in Design
• Overall volume of the cell after fabrication is 35 ml
• Four wire arrangement to measure voltage drop
independently from power wiring
• Incorporated a spring to provide a uniform tension and
avoid any slackness due to expansion
• Effective off-centering mechanical design provides
additional room for wiring and thermocouples
• Three thermocouples to verify the uniformity of the fluid
temperature
• Electrical connection junctions are arranged on the cell for
flexibility in connections and handling
• Boundary induced errors are minimized
www.kostic.niu.edu
 Conclusion
• Designed and Fabricated a Hot-wire cell with
improvements
• Designed and Fabricated a Wheatstone bridge for
Hot-wire cell
• Optimized Data Acquisition Hardware
• Developed a LabVIEW Program for Measuring
Thermal Conductivity
• Calibrated the Apparatus with Standard Fluids

www.kostic.niu.edu
 Conclusion
• Bias Error is within 1.5 %
• Precision Error is within 2.5 %
• Total Uncertainty within 3.5 % at 95 %
Probability
• Enhancement in Thermal Conductivity with
Copper in Ethylene glycol is 13 %
• Enhancement in Thermal Conductivity with
Copper in Water is 16 %

www.kostic.niu.edu
 RECOMMENDATIONS
• The uncertainty analysis shows that the resistors are the
major contributors of error. This error can be reduced by
using very high precision resistors with extremely small
temperature coefficient of resistance.
• In the present study, temperature coefficient of resistance
was determined through calibration over limited temperature
range. Precise calibration under well controlled conditions
with a larger temperature range would be beneficial.
• At present, the resistances are manually measured. This
process can be automated in future.
• The data acquisition and LabVIEW® can be programmed to
evaluate curvature of temperature versus logarithmic-time
dependence (at initial heat-capacity and later convection
non-linear regions), and automate evaluation if linear range
relevant for thermal conductivity measurement.
• The hot-wire tension can be more accurately controlled
using a micrometer in place of the fixed calibration gauge.
www.kostic.niu.edu
 Acknowledgements

The authors acknowledge support by

National Science Foundation
(Grant No. CBET-0741078).

The authors are also grateful for help in mechanical design

and fabrication to Mr. Al Metzger, instrument maker and
technician supervisor at NIU.

www.kostic.niu.edu
Thank You

www.kostic.niu.edu