trasferencia de calor

Magaly rosada
Diego Bermdez
Johanna Vargas

One approach to parameter estimation was applied to
characterize the heat transfer in scraped surface heat
exchangers (SSHEs) specifically designed for the food
industry. It is difficult to apply the data available in the
literature for SSHEs, due to the specificity of each
product, heat treatment and geometrical
configuration, causing the thermal design of these
devices is critical.
There are generally two to four blades, which can be
disposed longitudinally in the wall of the rotor
along the entire length of the heat exchanger.
Alternatively, short blades can be arranged in pairs
and offset by 180 with respect to the rotor axis, in
the presence or absence of a certain degree of
overlap. These two configurations are called direct
and alternating leaves, respectively
Although SSHEs are frequently used in industrial
applications
such as in the dairy and food industry, the scientific
literature on this topic contains some gaps, including
thermal design of these devices
although the numerical approach has led to some critical
results Most SSHEs studies are based on experimental
investigations of the forms of single-phase heat transfer
and two phase.
This is done assuming that the external resistance
is smaller q internal
Heat transfer to the correlation of the internal side,
in a manner monomial is generally adopted by the
dependence of the Nusselt number to the Reynolds
number of rotation and the Prandtl
while the dependence of the axial Reynolds
number is generally disregarded.
Uncertainty analyzes, provides information about
both the problem by allowing an assessment of the
quality and robustness of the correlations resulting
heat transfer.
For any heat exchanger of parallel flow, the transfer
coefficient Ai internal surface area can be obtained
from the equation:
where T is the logarithmic mean temperature difference and Q
is the heat flow rate.
In Eq. (1), the heat flow rate Q can be recovered by
determining
the energy balance of the secondary fluid stream
flowing in the
side as follows:
jo where expresses the outer side of the fluid specific enthalpy
evaluated in sections of output and input.
The overall heat transfer coefficient is related to the product
side and the heat transfer coefficients for convection of
secondary fluids by the following equation:
where hi is the heat transfer coefficient by convection flowing
in the tube with the rotor and Ho is the convection coefficient
of heat transfer of the fluid stream flowing on the side, Ao and
Ai are the areas of product transfer surface and inner and outer
heat Rw is the thermal resistance of the wall separating the
fluid, expressed as follows:
where kW is the thermal conductivity of the wall
material and L is the length of the heat exchanger.
The thermal resistance Rw of the wall can be
assumed to be known and constant for a given heat
exchanger under running conditions.
Using bore exchanger tube as the characteristic
length, inner side Nusselt number is expressed in
the following straightforward manner:
where k is the product (fluid in the inner side) thermal
conductivity.
Regarding SSHEs, under the hypothesis of a single
phase flow, the heat transfer coefficient of the product
side can be assumed to be correlated in terms of the
Nusselt number as follows:
where the Reynolds number is defined as rotation Rer
follows:
In Eq. (7), N is the rotational speed of viscosity and
density dynamics respectively.
This implies that the following function must be
minimized by using the least squares approach habitual :
where M is the number of measurements made for a mass
flow rate of the fluid by varying Rr and Pr, with resultant
UCalc expressed as follows:
Then, the parameter estimation method applied to the heat
transfer SSHEs characterization results from the minimization of
the equation given by S function. (8) assuming Rer and Pr as the
independent variables; C, , , and Ho as the unknown variables,
and all other quantities and geometrical properties as known.
The coefficients are defined with respect to the generic
parameter Pi in a dimensionless form as:
where Pi represents the unknown variables, C, , , and Ho, and U is
the overall heat transfer coefficient expressed as a function of the
independent variables and PR Rer. Following this approach, once the
best-fit curve P parameters are determined, the standard errors of
parameters RP
are given by:
where J is the Jacobian matrix of the target variable is the
UCalc function.
And 2 U is the residual variance:
where M is the number of measurements and z is the number of
parameters to be fitted.
In order to express the reliability of the parameter estimates
and compare the relative accuracy of the various estimates of the
parameters
confidence interval of 95%, the coefficient of variation,
CV is generally used with respect to the parameter Pi, defined as follows:
The sensitivity and uncertainty analysis applied to both synthetic
and experimental data for a coaxial SSHE.
It must be stressed that the approach presented here is based on the
assumption that the heat transfer coefficient on the outside varies
with the mass flow rate of secondary fluid is optimized for apparatus
in which a convection one phase is present in the secondary side
fluid.

The convective heat transfer coefcient of the uid
stream owing in the jacket side, namely h, was rst
varied over the range 1000- 4000 W/m2K, which is
representative of SSHEs in which the secondary uid
ows under the turbulent regime, as often happens
when water is used as the service uid.
The aim of the present investigation was to enable the
robust estimation of the heat transfer correlation for the
product side Nusselt number in an SSHE by assuming that
the values of the external side heat transfer coefficient are
unknown, as well. The estimation procedure presented
here did not require any hypothesis about the functional
dependence of the external side heat transfer coefficient, as
is generally assumed in modified Wilson plot approaches.
The procedure was validated through its application to
both synthetic and experimental data acquired from a
coaxial alternate blade SSHE pilot plant planned to treat
highly viscous fluid foods. The parameter estimation
procedure was optimized for sensitivity and uncertainty
analysis, which provided considerable insight into the
problem.