252865 Concepts Rendered False by Advances In Understanding and Obsolete by Advances In Computation
The formulation and adoption of useful concepts and generalizations, and in many instances unique ones, have helped chemical engineering to flourish as an academic subject and as a profession. The broadest ones are illustrated by unit operations, unit processes, and transport phenomena. Narrower ones are illustrated by the equilibrium stage, the perfectly mixed continuous reactor, and the Colburn j-factor. These latter concepts have resulted from observations in plant operations and in the laboratory as well as from theoretical analyses and flashes of insight. All in all, they constitute an essential element of both education and practice in chemical engineering.
With the passage of time, most of the broad concepts have been replaced, modified, or supplemented. Many of the specific ones have been found to be false or unnecessary yet remain in textbooks and computer packages out of inertia and/or out of misdirected respect for those who originated them. The objective of this presentation is to identify some of the obsolete and/or false ones. Within this restricted space, identifications can only be illustrative, so an attempt has been made to choose examples that are diverse in subject and in the nature of the shortcomings. The results of the advances by virtue of computation will be shown graphically in the presentation. It is hoped that these examples will inspire each of you to question the continued viability of the concepts and idealizations in your personal professional portfolio, including the books you depend upon.
The formulation of each of the concepts discussed herein depends upon one or more ingenious idealizations and/or simplifications. With time some of these have become unnecessary because of advances in computer hardware and software, and some have been proven to be false by advances in understanding. Their identification is an essential part of the process of re-examination of a concept, and often the critical one.
The concept of reaction engineering was a great advance over chemical kinetics, which was
originally taught as a subtopic in physical chemistry. It continues to have vitality but has accumulated a number of false concepts, a few examples of which are noted here.
This is a misleading, inaccurate, and totally unnecessary concept. It should be excised from our literature and replaced by realistic fluid-mechanical models.
Space velocity and space time
These concepts invoke a Lagrangian framework and thereby the shortcomings of plug flow.
They too should be excised.
Perfect radial mixing in tubular flow
This is a more accurate verbal description of the behavior resulting from the postulate of plug flow in that it can be considered as an asymptote for Pr → 0 and Sc→ 0, but it does not ameliorate the errors in the model.
Fully developed flow with no radial mixing
This is the counterpart of perfect radial mixing, that is, it corresponds to Pr→ ∞ and Sc→ ∞, and as such is physically conceivable. It is useful as a lower-bound for the conversion and as an upper-bound for the length of the reactor, which are conservative predictions and thereby good engineering as contrasted with the predictions of perfect radial mixing which are non-conservative and thereby poor engineering.
Turbulent tubular flow
Chemical conversions rarely occur in the turbulent regime because the reactor is then almost
certainly of excessive length. Plug flow is not approached; the velocity remains nearly parabolic.
The initiation of a reaction over the entire cross-section at the inlet is implied in most models
although it is impossible to achieve physically. This idealization is difficult to replace, because
the alternatives involve developing flow and thereby an increase of an order of magnitude in
complexity of the numerical modeling. Until such modeling is included in computer
packages, the affects of this non-conservative idealization should not be ignored.
The expression of reaction rates in terms of molecular concentrations rather than activities
betrays the principles of chemical engineering thermodynamics. To brag about its rigor and
generality as compared to that of mechanical engineering and chemistry and then cling
to this erroneous concept is untenable. It is difficult to re-do rate constants in terms of
concentrations because that requires a knowledge of the pressure, temperature, and
composition at which the measurements were made. The error in the predictions is serious
only if the conditions differ significantly from those of the measurements, but the time
has come for the editors of chemical engineering publications to establish a new standard.
As shown definitively in an article by Mayer and Stowe, models of the type proposed by
Hougen and Watson for chemical conversions in flow through a bed of catalyst particles
introduce more empirical coefficients than can be justified by the precision and extent of the
experimental data. It follows that the resulting models have no physical significance. This
approach should have been abandoned long ago. This is an example of the misuse of
numerical computation. This listing could go on and on and on; the number of obsolete and
false concepts and the unnecessary idealizations and simplifications in reaction engineering is
almost endless, but some space must be reserved for other topics.
Combustion has progressed from art to an engineering science by virtue of advances in
computation and in the determination of free-radical mechanisms. It should be a sub-topic,
such as Separations, within chemical engineering but because of the failure to encompass it
in reaction engineering it has become a separate profession. The probable reason for such
short-sightedness is that combustion occurs primarily in unconfined flow whereas chemical
engineers are more comfortable with tubular flow. Two false concepts related to the
modeling of combustion deserve mention. First, solids, liquids, and gases do not burn – only
free radicals; therefore combustion must be preceded by pyrolysis. Second, the concentration
of free radicals never attains a stationary state. Global modeling should be abandoned.
Molecular separations are, along with reaction engineering, unique to chemical engineering.
There is space to mention only one ingenious but now obsolete concept, namely, the ideal-
stage. The many inherent idealizations, most of which are now avoidable owing to advances
in computation, should not go unmentioned.
Chemical engineers have more or less ceded leadership in heat transfer to mechanical
engineers although heat exchange remains essential in chemical processing and in
petroleum refining. Space allows the mention of only two obsolete concepts and one
Correlations in the form of products of powers of Re and Pr
A power-dependence actually occurs, if at all, only in an asymptotic sense. As an example,
the Colburn analogy is functionally in error in every respect and, as a consequence,
seriously in error numerically. The formulation and use of far more accurate correlating
equations in the form of power-means of the asymptotes has been encouraged and abetted
by advances in computation.
The hyperbolic equation of conduction
This model is totally false. In the instance of a pressure wave generated by impulsive
heating, such as thunder by lightning, the behavior can be predicted exactly by
taking the compressibility and the conservation of momentum into account. The
computational investment is however high owing to stiffness.
The enhancement of convection by an energetic reaction
This effect, which is completely overlooked in all textbooks has been resolved by differential
models and finite-difference computations.
This subject is shared with mechanical engineers, civil engineers, and others but chemical
engineers remain the pace-setters in special topics such as non-Newtonian fluids, packed
beds, and fluidized beds. Obsolete concepts are illustrated by the following examples.
The power-law fluid
Such behavior is simply a necessary artifact of the transition between pseudoplastic and
The orifice coefficient
The numerical expression derived for a sharp-edged orifice on the basis of a free streamline is a misleading approximation. The coefficient, including its asymptotic value, varies with Do /D and Re, and should be computed numerically.
The Ergun equation
The attribution of the Re2 term of the Ergun equation to turbulent rather than inertial flow is
an example of a pervasive misinterpretation.
DNS (direct numerical simulation) and LES (large eddy simulation)
These methodologies exemplify self-defeating limitations. DNS has proven to be viable only
for the hypothetical case of a parallel-plate channel and for Reynolds numbers barely above
that for fully developed turbulence. LES has proven to have the same geometrical limitation
and to require very empirical supplementation.
The analogy between heat and mass transfer
The substitution of Sh and Sc for Nu and Pr has been shown both experimentally and by
Lagrangian direct numerical simulationto be invalid.
The Lewis number
In a session in his honor at the 57th AIChEAnnual Meeting, Warren K. Lewis jokingly
complained that the dimensionless group named in his honor commemorates one of his
worst mistakes, namely the presumption that because that grouping happened to be
nearly equal to unity in some of his experiments with water vapor in air it was unity for all
conditions and systems.