Sie sind auf Seite 1von 9

Infrared Physics & Technology 45 (2004) 93101

www.elsevier.com/locate/infrared

Rapid, non-destructive and non-contact inspection


of solid foods by means of photothermal radiometry;
thermal eusivity and initial heating coecient
a,b a,*
A. Gijsbertsen , D. Bicanic , J.L.W. Gielen c, M. Chirtoc d

a
Laser Photoacoustic Laboratory, Biophysics Division, Department of Agrotechnology and Food Sciences, Wageningen University,
Dreijenlaan 3, 6703 HA Wageningen, The Netherlands
b
Laser Center and Department of Physical Chemistry, Vrije Universiteit Amsterdam, De Boelelaan 1083, 1081 HV Amsterdam,
The Netherlands
c
Department of Mathematical and Statistical Methods, Biometris, Wageningen University, Bornsesteeg 47, 6708 PD Wageningen,
The Netherlands
d
Institute of Isotopes and Molecular Technology, P.O. Box 700, 6400 R Cluj-Napoca 5, Romania
Received 9 January 2003

Abstract

CO2 -laser photothermal radiometry (PTR) was demonstrated to be suitable for the non-destructive and non-contact
characterization (both optical and thermal) of solid phase agricultural commodities (fresh vegetables, fruits) and
confectionery products (candy). Proper interpretation
p
of PTR signals enable one to calculate two parameters, i.e. the
well known thermal eusivity e (e kqcp , where k and qcp are the thermal conductivity and the volume specic heat,
respectively) and a newly introduced physical quantity termed initial heating coecient chi (v b=qcp , b is the
absorption coecient). Obtained values for e are in a good agreement with data reported in the literature. PTR enables
one to rapidly determine e via a single measurement. As opposed to this, the knowledge of two out of three thermo-
physical parameters (thermal diusivity, thermal conductivity and volume specic heat) is a condition sine qua non for
determining eusivity in the conventional manner.
 2003 Elsevier B.V. All rights reserved.

1. Introduction of (temperature dependent) thermophysical pa-


rameters. Properties such as volume specic heat
There is a continuous and a widespread demand (qcp ), thermal conductivity (k), diusivity (a) and
in the food industry for accurate numerical values thermal eusivity (e) are required to predict heat
transfer rates needed for the design, operation and
control of numerous processes concerning foods
*
Corresponding author. Tel.: +31-317-484954; fax: +31-317- and are also related to the quality of foods [1]. In
482725. principle, the thermal properties of foods can be
E-mail address: dane.bicanic@wur.nl (D. Bicanic). predicted if their composition, porosity and

1350-4495/$ - see front matter  2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.infrared.2003.07.003
94 A. Gijsbertsen et al. / Infrared Physics & Technology 45 (2004) 93101

temperature are known [2]. The earlier heat In the past, argon-laser based PTR was used to
transfer analysis of the heating and cooling of food monitor qualitative changes induced in margarine
products was actually oversimplied, as it relied on exposed to UV light [19] and to characterize the
constant, uniform values of thermal properties. drying of cheese [20], while infrared ash ther-
Modern techniques are capable of dealing with mography was used to monitor the rmness of
thermal properties that alter with time, tempera- fresh apples and the ripeness of raw tomatoes [21].
ture and location during the treatment of foods [3 The main drawback in using Ar-laser radiation for
7]. The two major reasons for the general lack of the heating of biological tissue is (partial) trans-
data concerning the thermophysical properties of lucence of such samples; this induces scattering
foods are the complex nature and the ever in- within the sample and generally a weak signal,
creasing number of available food products. The which complicates greatly the quantitative inter-
fact that conventional characterization techniques pretation. The problem of incomplete opacity is
are destructive, time consuming and in addition less serious if CO2 laser is used for the excitation.
often require substantial quantities of test sample However, in this case there is the problem in that
for analysis, is stimulating research attempts to- the excitation wavelength (10.6 lm) and the
wards the development of new methods. wavelength corresponding to the maximum of
Photoacoustic (PA), photothermal (PT) and re- spectral emittance (at room temperature) are
lated methods are emerging as novel techniques for practically in the same wavelength region. This
the measurement of thermal and optical properties might cause problems when trying to discriminate
of samples in practically any physical condition between the incident laser radiation and the ther-
[8,9]. Common to these techniques is the detection mal radiation emitted by the sample under inves-
of heat (or the eects of this heat) generated as a tigation.
result of the samples absorption of periodically This article describes the application of the
modulated (or pulsed) incident radiation. CO2 -laser for quantitative PTR studies of solid
So far, some PT methods have been used for the fresh agricultural commodities (vegetables and
thermal characterization of foods (solids, liquids, fruits) and confectionary products. The thermal
pasty and gels) and a few biological materials. Two eusivity (e) and initial heating coecient (v) of
recent examples are (i) the concept of a thermal these samples were obtained directly from the
wave resonant cavity (compatible with non- phase and the amplitude of the PTR signal.
aggressive liquids) [10,11] that has proved useful in
highly accurate measurements of the thermal dif-
fusivity a and (ii) the inverse photopyroelectric
technique (IPPE), a simple and very ecient ap- 2. Theory
proach towards directly obtaining thermal eu-
sivity e [12,13]. The IPPE as well as its (Standard) Essentially, PTR measures the emittance of the
counterpart SPPE, require intimate thermal con- sample, that depends on the surface temperature
tact between the detector and the sample under (StefanBoltzmann law). In the experiments de-
investigation. Such a requirement is easily met for scribed below, the sample is periodically heated
(semi-) liquids and gels, but not for solid samples due to the absorption of laser radiation. Since,
that are usually investigated by photothermal relative to the ambient temperature, the periodical
beam deection (PTBD), photoacoustic (PA) change of the surface temperature is rather small
spectroscopy and photothermal radiometry (PTR) (of the order of 7 mK at 100 Hz), there is a direct
[14,15]. Both PTBD [16] and PA [17] enable the proportionality between the latter quantity and the
experimentalist to obtain the thermal diusivity emittance and thus the PTR signal as well.
(a). In addition, Balderas-L opez et. al. applied the All test samples studied here are thermally and
temperature rise technique to determine the vol- optically thick, which means that their thickness
ume specic heat (qcp ) and then used this value to (L) is much larger thanqthe
penetration depth of
2a
calculate e [18]. the thermal wave (l x , where a is the thermal
A. Gijsbertsen et al. / Infrared Physics & Technology 45 (2004) 93101 95

diusivity, and x is the angular frequency) and the 


A
penetration depth of the optical wave (1=b). Since Te x; s F s exp  bx
only a part of the samples area illuminated by a s  ab2
r  r
modulated laser beam is actually projected on the a s
b exp  x : 4
active element of the detector, the use of a one s a
dimensional model is justied. Using the energy
Since the emitted radiation originates from the
conservation law, the following partial dierential
samples surface (water present in the test sample
equation can be derived [17].
has an absorption peak in the detection band-
Tt x; t aTxx x; t A f t expbx; 1 width), one can substitute x 0 in Eq. (4), which
then reduces to:
with:
Te 0; s GsF s; with
 bI0  r
A qcp
initial temperature gradient ( vI0 ) A a
[K s1 ] Gs 1  b : 5
s  ab2 s
a k
qcp
thermal diusivity [m2 s1 ]
T temperature (relative to ambient temper- In Eq. (5), Gs is the transfer function, while F s
ature) [K] is the normalized input signal (i.e. laser intensity
I0 incident laser intensity (unmodulated) It divided by maximum intensity I0 ). As to Gs,
[W m2 ] it can be written in a more simplied form:
b absorption coecient per unit length A
[m1 ]. Gs p : 6
s b as

The unmodulated incident laser intensity I0 Assuming sinusoidal heating, the phase and am-
(It I0 f t) used in Eq. (1) is already corrected plitude response can be found after substituting
for reection loss at the surface of the sample with s ix, where x [rad s1 ] is the (angular) modula-
reectivity r. The incident intensity Iin and the in- tion frequency:
tensity I0 entering the sample are related through: A
Gix p : 7
I0 1  rIin : 2 ix b aix
Assuming that (i) the heat exchange between The amplitude (dT ) and phase (/) can now be
the sample and the surrounding air is negligible calculated by taking the absolute value and the
compared to the heat ow due to conduction into argument of Gix:
the sample and (ii) the initial temperature is uni- q
2 2
form, the boundary conditions (bc) and the initial dT jGixj RGix IGix ; 8
condition (ic) read as:  
IGix
bc1 : Tx 0; t 0; / argGix arctan : 9
RGix
bc2 : T 1; t 0;
ic : T x; 0 0: Substituting A in Eq. (7) and thus obtained Gix
in Eqs. (8) and (9), gives the following expressions
The Laplace transform: L : T x; t ! Te x; s and for the amplitude and phase:
L : f t ! F s, changes Eq. (1) into:
bI0
dT q
p ; 10
o2 Te x; s s e A
 T x; s  F s expbx: 3 qcp xx b2 a b 2ax
ox2 a a
Solving this equation by using the boundary con- p !
2x
ditions given above, one obtains the temperature /  arctan 1 p : 11
function Te x; s in the Laplace domain: b a
96 A. Gijsbertsen et al. / Infrared Physics & Technology 45 (2004) 93101

Thispequation
can be rewritten in terms of I0 bc
e kqcp [J m2 K1 s1=2 ] and v b=qcp H ix
qcp b c
[m2 KJ1 ]. The rst quantity is the well known p p
thermal eusivity, the latter is termed the initial ix ab c
 p p p p p :
heating coecient. Mathematically, this latter is ix ac ix ix ab
the time derivative of the surface (x 0) temper- 18
ature, when the laser initially (t 0) strikes the
sample. Both the phase dierence (h) between the modu-
lated incident laser radiation and the radiation
Tt x 0; t 0 vI0 K s1 : 12 emitted by the sample and the amplitude (dM) of
the radiation emitted by the sample can be calcu-
Eqs. (10) and (11) now become: lated by taking, respectively, the argument and the
vI0 absolute value of H ix:
dT q
p ; 13
xx e2 v2 ev 2x h argH ix; 19

p ! dM jH ixj: 20
2x
/  arctan 1 : 14 Eqs. (19) and (20) are not only cumbersome but
ev
in addition require the availability of one addi-
tional parameter (c) that must be estimated from
From Eqs. (13) and (14), one can calculate e and v
measured data. An attempt was made to use this
directly from the measured phase and amplitude.
improved model in order to t the frequency de-
p
2x pendent measured data for agar and to estimate
v ; 15 the model parameters i.e. b, c, a and qcp . More-
etan/  1
over, to reduce the number of model parameters
p one also tried to set b c thereby making it pos-
2I 0
e p : 16 sible to estimate e and v. Neither attempt, how-
2
xtan / 1dT ever, resulted in a signicant improvement in the
To some extent, a shortcoming of the model t to the frequency dependent measurement data
above is the fact that it relies on the amplitude and of the amplitude and phase dierence for agar.
phase of the time dependent surface temperature. Because of this and the simplicity of the surface
Such a model is only valid if the mean absorption temperature model, Eqs. (15) and (16) were used to
coecient (c) of the sample in the detection perform the calculations, the outcome of which is
bandwidth is much larger than that (b) at the laser described in this paper.
wavelength. If this is not the case, then the model
discussed above has to be improved by multiplying
the temperature function in Eq. (5) by ecx and 3. Materials and methods
then integrating the temperature spatially to ob-
tain the time dependent emitted radiation Ms in The experimental setup used for the PTR
the Laplace domain: measurements is shown in Fig. 1. The excitation
Z 1 source is a c.w. line tuneable waveguide CO2 -laser
A (A) tuned to the 10P20 line (10.55 lm), providing
Ms F s exp  bx
s  ab2 0 as much as 3 W output power. At this wavelength,
r  r
a s the absorption of water exhibits a broad spectral
b exp  x ecx dx: 17 peak. Laser radiation, modulated by a home-made
s a
mechanical chopper (B), was reected at plane
The transfer function H s can be calculated after gold coated mirrors (C, G, H) and directed onto
integration; substituting s ix gives: the sample (K). A beam director (D) inserted be-
A. Gijsbertsen et al. / Infrared Physics & Technology 45 (2004) 93101 97

error for e and v. The amplitude depends on r as


2
1  r (assuming that the reectivity at the laser
and at the detection wavelength is the same), while
the phase lag is surface reectivity independent.
Furthermore throughout all the measurements
thermal imaging plates where used to control the
spatial distribution of the laser beam.
The thermal radiation emitted by the absorbing
sample was collected by the same parabolic re-
ector (J) and then focussed on a high speed, li-
quid nitrogen cooled Mercury Cadmium Telluride
detector (L) from Infrared Associates (MCT-
Fig. 1. Experimental set-up used for PTR studies in IR. 13-1.0, ZnSe wedge) mounted on a scissor jack and
positioned in the same vertical plane as the sample.
To prevent the forward scattered laser radiation
tween C and G enables one to identify (F) and to from hitting the detector and causing undesirable
measure (E), the power of the incident laser signals, a LiF lter (M) was used right in front of
emission. the detectors entrance window. The LiF lter is
For actual PTR measurements, CO2 laser ra- open for wavelengths shorter than 8 lm and
diation was rst allowed to pass through an AR/ practically opaque for wavelengths longer than 9
AR coated ZnSe lens (I, Laser Power Optics, lm. The detector sees radiation the wavelength
f 254 mm, 24:5 mm) and a circular hole of which exceeds 5 lm. This means that the actual
( 5 mm), drilled o-axis in a Leybold 389 24 B detection takes place around the dominant ab-
aluminium parabolic (f 300 mm, 420 mm) sorption peak of water at 6.08 lm, implying that
reector (J) before striking the sample (K). The the radiation to be detected originates from the
sample was mounted in a holder (positioned at very surface of the sample. This justies the use of
2f 600 mm from J) that itself was axed on a the above-discussed model (Eqs. (15) and (16)),
xyz stage. The choice of the focal length and the which assumes that the detector sees solely the
location of ZnSe lens were such that the diameter surface and not the inside of the test sample.
of the laser beam on the sample was about 8 mm. The PTR signal from the detector was pre-am-
In order to prevent damage of the sample by in- plied and then fed into a two phase 3961B Ithaco
cident radiation, the output power of the laser was lock-in amplier. The alignment of the set-up was
reduced by adjusting the discharge conditions. For checked by measuring the lock-in signal obtained
the studies on agricultural commodities, the laser either without the sample or from a strongly re-
power measured directly behind the chopper was ecting sample; in both cases the PTR signal was
between 150 and 600 mW corresponding to 0.20.8 approximately zero (<3% of signal for agar gel). In
W cm2 intensity at the sample. The spatial dis- addition, a TiNx surface absorbing reference
tribution of the laser beam was controlled using a sample (dimensions 50 mm 50 mm 5 mm) of
thermal imaging plate. quarts coated with 1.1 lm thick layer of TiNx
At the excitation wavelength, the surface re- (DIN 50992-2, kindly provided by Phototherm Dr.
ectivity from biological tissue varies from one Petry GmbH) was used to test the set-up. For such
sample to the other, but it is generally weak. a reference sample the phase dierence should be
Typical values for (1  r) for biological samples )45 and independent of the modulation fre-
and water span the range extending from 0.95 to quency [24]. Indeed, with TiNx as a test sample the
0.99 [22,23] at wavelengths around 10 lm. In this phase dierence was shown constant over the en-
study it is assumed that the surface reectivity for tire frequency range, while p the
amplitude was di-
all samples studied here is equal to that of agar; rectly proportional to 1= f as predicted by
This assumption results in a 45% experimental theory. The result of a test measurement is shown
98 A. Gijsbertsen et al. / Infrared Physics & Technology 45 (2004) 93101

in Fig. 2. At a given modulation frequency, the radiation. The modulation frequency was then
PTR signal from the selected site was measured changed and step (iv) repeated; such a procedure
during a 20 s long interval. was performed for at least six dierent frequencies
The TiNx reference sample was also used to spanning the range between 10 and 100 Hz and
determine the instrumental phase (IP []). On the completed by the control measurement of laser
other hand, agar gel (1% agar, with thermal output power. The test sample was then removed
properties same as water [25]) and graphite from the holder, agar (or graphite) mounted in-
(Ringdor) were used to obtain the instrumental stead and IC determined again to make sure that it
constant (IC [V W1 K1 ]). While IP represents the did not change during the measurement. If, during
frequency independent oset between the phase a measurement, the signal did not reach a constant
read from the lock-in amplier (/li ) and the real level (although both IC and power P did not
(theoretical) phase (/), IC is a constant giving the change) but instead continued to increase slowly
relationship between the amplitude reading of the due to a drying process induced by heating, the
lock-in Ali [mV] (normalized to the incident power sample was slightly displaced prior to performing
P [mW] of the laser) and the amplitude of the the next measurement. In this manner dierent
surface temperature (dT ): regions on the sample were irradiated in each
/ /li  IP; 21 consecutive measurement. Such a drying eect
causing a drop of e with time (which generally
Ali induces an increase of amplitude) was minimized
dT : 22 by reducing the laser output power.
P IC
Once IC and IP were determined, PTR mea-
surements on various test samples were made. The
4. Results and discussion
measuring protocol included the following steps:
(i) mounting of the sample, (ii) measurement of the
To demonstrate the potential of PTR to obtain
laser power P , (iii) selection of the modulation
the thermal eusivity of solid foods in a simple and
frequency and (iv) recording (lock-in amplier) the
rapid manner, one has studied fresh fruits and
amplitude and the phase of the PTR signal for
vegetables (tomato, pear, apple etc.) and confec-
about 20 s after exposing the sample to the laser
tionary products; the results are summarized in
Table 1. The test samples were purchased at a local
supermarket. Frequency scans obtained from
4
lock-in amplitude [mV]

several test samples are plotted in Fig. 3.


3 The standard deviation for v reported in Table
2 1 refers to ve consecutive measurements at 12.8
1
Hz. The standard deviation for e was calculated
using the data obtained from measurements con-
0
0 0.05 0.1 0.15 0.2 0.25 0.3 ducted at ve dierent frequencies at least. Based
1/sqrt(f) [1/sqrt(Hz)] on measurements on agar gel and graphite (these
80 reference samples have well known properties), a
lock-in phase [deg]

60
typical accuracy for e attained in PTR measure-
ments is estimated to be 5%.
40
It was found in this study that, dierent from
20
what was expected, all test samples exhibited a
0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 slight decrease of v (depending on the sample this
1/sqrt(f) [1/sqrt(Hz)] decrease varies between 10% and 20%) for in-
Fig. 2. The amplitude (top) and the phase (bottom) of the PTR creasing modulation frequency. The initial heating
signal obtained from the TiNx reference sample used to test the coecient v is directly proportional to b via
experimental set-up. v qcbp . The reader is reminded that values for v
A. Gijsbertsen et al. / Infrared Physics & Technology 45 (2004) 93101 99

Table 1
Thermal eusivity and initial heating coecient of foods at 293 K acquired by PTR
Sample v (103 m2 KJ1 at 12.8 Hz) e (103 J m2 K1 s1=2 )
Chocolate (pure) 19.6 0.5 0.83 0.03
Chocolate (pure, melted) 21.7 0.3 0.622 0.005
Peppermint 36.0 0.6 0.562 0.012
Tomato (normal) 26.5 0.8 1.35 0.03
Tomato (fresh) 28.2 0.9 1.31 0.05
Butter bean 12.4 1.2 0.42 0.05
Apple (Granny Smith) 17.5 0.6 1.20 0.05
Apple (Jonagold, 12.8 Hz) 23 4 1.32 0.03
Apple (Jonagold, 98 Hz) 14.0 0.5 (at 98 Hz) 1.07 0.06
Pear (D. du Commice) 18.3 0.1 1.23 0.03

25
chocolate (pure)
simultaneous decrease of polymer structures on
20
peppermint
pear approaching the surface. The net result of such a
agar gel
phenomenon is a decrease of the absorption coe-
A [mV]

15

10 cient (b) (at wavelengths associated with polymer


5 structures) with increasing modulation frequency.
0 Similar phenomenon may also take place in the
10 20 30 40 50 60 70 80 90 100
f [Hz] waterlike samples such as those investigated in this
55 study.
50 For practically all investigated samples e was
phase [deg]

45 found to be independent of modulation frequency.


40 The only exception was a Jonagold apple, that
35 exhibited relatively large dierences in both e and
30
v values. Presumably, this is due to dierences in
10 20 30 40 50 60 70 80 90 100
f [Hz] the properties of the skin and lower layers. It
might be that at 98 Hz it is only the skin that is
Fig. 3. The amplitude (top) and the phase (bottom) of the PTR
signal obtained from four dierent samples plotted versus being probed, while the underlaying tissue plays a
modulation frequency. The IC and IP are 570 V W1 K1 and role at 12.8 Hz. The large standard deviation for v
106 respectively. The incident laser power P at 10P20 measured can probably be ascribed to the variations of the
at E (Fig. 2) was 560 mW for confectionary products as com- skin thickness.
pared to 600 mW for agar gel and fruits.
From Table 1 one concludes that e correlates
well with the moisture content of the sample;
given in Table 1 refer to a measurement at 12.8 Hz butterbean that is rather dry has a low e. Clearly,
(at this lowest attainable frequency in our mea- the higher the moisture content of the sample, the
surements, the thermal diusion length is maxi- closer the value of e to that of water, i.e. 1585
mal). A possible explanation for the observed [J m2 K1 s1=2 ].
dependence of v on f can be attempted using data The use of PTR for the measurement of eu-
of Haas et al. [26], who conducted a PA study of sivity is not limited to studies of fresh fruits and
liquid water, heavy water and alcohol in the near vegetables. Other samples featuring low reection
infrared. They concluded that these samples (ex- or no reection at all can also be investigated, as
hibiting the association to polymerlike structures demonstrated here (Table 1) by studies on choco-
through hydrogen bonds) reveal a structural late (hard and melted as a result of interaction
modication from the interior of the liquid to- with laser radiation) and peppermint candy. Val-
wards the liquidgas boundary. This was ex- ues of e obtained by PTR for chocolate are close to
plained [26] by an increase of monomers with a those from the literature [27]. While the eusivity
100 A. Gijsbertsen et al. / Infrared Physics & Technology 45 (2004) 93101

of melted chocolate was found to be slightly larger van Asselt, M. Schimmel and O. Ganguli (Wa-
than that of solid chocolate, the trend in the values geningen) for their help received during this study.
of v is just the opposite.

References
5. Conclusion
[1] H.F.T. Meert, History, aims, results and future of
To the best of our knowledge, this paper de- thermophysical properties work within COST 90, in: R.
scribes the rst application of room temperature Jowitt et al. (Eds.), Physical Properties of Foods, Applied
CO2 laser PTR for studies of foods. The spectral Science Publishers, London, 1983.
[2] A.D. Brailsford, K.G. Major, Brit. J. Appl. Phys. 15 (1964)
overlap of incident and emitted radiation (the 313319.
major experimental diculty normally expected in [3] A. Mandelis (Ed.), Non-destructive Evaluation, volume II
this kind of studies) could be eectively suppressed of the series: Progress in Photothermal and Photoacoustic
by inserting a LiF lter in front of the detector. Science and Technology, Prentice Hall, Englewood Clis,
The results of this study conrm the suitability 1994.
[4] R. Santos, L.C.M. Miranda, J. Appl. Phys. 52 (6) (1981)
of room temperature CO2 PTR for the precise 41944198.
(5%) assessment of thermal eusivity (e) and the [5] R.D. Tom, E.P. Ohara, D. Benin, J. Appl. Phys. 53 (8)
initial heating coecient (v) of solid foods. The (1982) 53925400.
eusivity values obtained by PTR agree with cal- [6] W.P. Leung, A.C. Tam, J. Appl. Phys. 56 (1) (1984) 153
culated values reported in the literature [27]. The 161, Erratum 58 (2), 1087.
[7] S. Prahl, Pulsed photothermal radiometry of inhomoge-
information obtained from a PTR measurement is neous tissue, in: A. Mandelis, P. Hess (Ed.), Progress in
related only to the properties of the surface layer Photothermal and Photoacoustic Science and Technology,
and near-surface layer. The PTR approach de- vol. III, 1997, pp. 416438.
scribed here requires a calibration with an opti- [8] P. Dorr, C. Gruss, J. Appl. Phys. 89 (2001) 78887894.
cally opaque (b ! 1 at excitation wavelength) [9] B. Majaron, W. Verkruysse, B.S. Tannenbaum, T.S.
Milner, J.S. Nelson, Phys. Med. Biol. 47 (2002) 1929
reference sample having well known thermal 1946.
properties. [10] J.A. Balderas L opez, A. Mandelis, J.A. Garcia, Anal. Sci.
In conclusion, CO2 laser based PTR is a new 17 (2001) 519522.
and practical technique for the rapid character- [11] J. Shen, A. Mandelis, Rev. Sci. Instr. 66 (1995) 49995005.
ization of foods. Being by virtue of its operational [12] D. Dadarlat, J. Gibkes, D. Bicanic, A. Pasca, J. Food Eng.
30 (1996) 155162.
principle, non-destructive and non-contact, PTR [13] D. Dadarlat, D. Bicanic, H. Visser, F. Mercuri, A.
can in addition also be used for real time moni- Frandas, JAOCS 72 (1995) 273279.
toring of e and v; practical examples include fol- [14] S.O. Kanstad, P.E. Nordal, Phys. Scripta 20 (1979) 659
lowing processes in which e or v change; i.e. 662.
drying, heating, cooling, melting etc. Another po- [15] R.E. Imhof, B. Zhang, D.J.S. Birch, Photothermal radi-
ometry for nondestructive evaluation, in: A. Mandelis
tential application is the assessment of food (Ed.), Progress in Photothermal and Photoacoustic Science
spoilage [28]. and Technology, PTR Prentice Hall, Englewood Clis,
The technique can be adapted to allow studies 1994.
on liquid and pasty samples as well. Finally, the [16] S.M. Brown, D. Bicanic, K. van Asselt, J. Food Eng. 28
feasibility of the PTR method for measurements of (1996) 221223.
[17] A. Rosencwaig, Photoacoustics and Photoacoustic Spec-
e at elevated temperatures is anticipated. troscopy, John Wiley & Sons, New York, 1980.
[18] J.A. Balderas L opez, S.A. Thomas, H. Vargas, V. Olalde-
Portugal, R. Baquero, I. Delgadillo, J.M. Yan ~ ez-Limon,
Acknowledgements J.J. Alvarado-Gil, J. Frias-Hernandez, L. Scheinvar, C.
Falconi, M.D. Silva, L.C.M. Miranda, Forest Prod. J. 46
(4) (1996) 8489.
The authors express their gratitude to H.G. [19] J. Gibkes, D. Bicanic, J. Cozijnsen, R. Frankhuizen, R.
Walter (Jena), J.C. Krapez (Chatillon), W. Mastop Koehorst, O. Doka, H. Jalink, Detection of changes
(FOM Rijnhuizen), R. Wolthuis (Rotterdam), K. induced in margarine exposed to air and to low level UV
A. Gijsbertsen et al. / Infrared Physics & Technology 45 (2004) 93101 101

radiation: an intercomparison study, in: F. Scudieri, M. [23] Available from <http://www.thermoworks.com/products/


Bertolotti (Eds.), Photoacoustic and Photothermal Phe- ir/mtfs.html>, May 2003.
nomena: Proceedings of the 10th Topical International [24] T. Mangold, Dynamische Thermography mit einer Schn-
Conference on Photoacoustic and Photothermal Phenom- ellen IR-Array-Kamera, Dissertation zur Erlangung des
ena, AIP Conference Proceedings 463, American Institute Grades des Doctors der In-genieurwissenschaften, Univer-
of Physics Press, Woodburry, NY, 1999, pp. 670672. sitat des Saarlandes, Saarbr
ucken, Germany, 2000 (in
[20] B. Bein, J. Gibkes, I. Delgadillo-Holtfort, D. Bicanic, J. German).
Pelzl, Moisture uptake from humid atmosphere analysed [25] I.A. van Haneghem, Een niet stationaire naaldmethode
by infrared radiometry, paper presented at Quantitative (warmtegelei-ding, warmtecapaciteit, contactweerstand),
Infrared Thermography (QIRT) Conference, Reims, 1721 PhD thesis, Wageningen University, Wageningen, The
July 2000, private communication. Netherlands, 1981 (in Dutch).
[21] S. Oermann, D. Bicanic, J.C. Krapez, D. Balageas, E. [26] U. Haas, Can. J. Phys. 64 (1985) 10631066.
Gerkema, M. Chirtoc, M. Egee, K. Keijzer, H. Jalink, [27] G.D. Hayes, Food Engineering Data Handbook, Long-
Instrum. Sci. Technol. 26 (1998) 145156. man Scientic & Technical, Harlow, England, 1987.
[22] Ircon, Inc., UltimaxJR portable infrared thermometer [28] O. Krischer, H. Esdorn, VDI-Forschungsfeht 450 (1956)
model UX-400 instruction manual, Revision C April 1998. 2839.

Das könnte Ihnen auch gefallen