Argentum Solutions, Inc.
Sterling guidance on corrosion and materials degradation
| Potential-pH Diagrams |
 |
|
| Intelligent Tools |
 |
|
 |
|
 |
|
| Corrosion Calculator |
 |
|
| Corrosion Economics Estimator |
 |
|
|
David C. Silverman |
|
Laboratory Corrosion Testing Under Heat Transfer Conditions – a Critique The transfer of heat through a surface can affect corrosion through a combination of increased surface temperature and, especially in the case of natural convection, by inducing fluid motion where it otherwise might not exist. The Tutorial on CYLEXPERTTM and the Rotating Cylinder Electrode discussed using the rotating cylinder electrode for corrosion prediction in the presence of single phase fluid motion but in the absence of heat transfer (under isothermal conditions). Increasing the temperature of the fluid and using the rotating cylinder electrode or equivalent to assess corrosion under dynamic conditions does not duplicate the effect of heat transfer through a wall. The thermal boundary layer disappears when the specimen reaches the fluid temperature. The wall temperature is different than when heat transfer occurs to maintain the same temperature. Thermal effects on mass transfer are also absent when the temperature differential disappears. Proper testing of heat transfer effects on a corroding surface through which heat transfer occurs requires that heat transfer through that surface be included in the testing.The relative values of the Prandtl number and Schmidt number provide a good illustration of the relative effects. This figure )
shows how the three boundary layers might develop for liquids in which
the Prandtl number is greater than 1, the Schmidt number is much
greater than one, and heat and mass transfer both begin at the beginning of a
horizontal surface. The kinematic viscosity (diffusivity
of momentum) is greater than both the thermal diffusivity and the
mass diffusivity in this fluid. The result is that both the thermal
and mass transfer boundary layers lie within the momentum boundary layer.
In this picture, the mass transfer boundary layer lies within the thermal boundary
indicating that the free stream concentration is reached at other than
the free stream temperature. The free stream concentration is reached
at a distance shorter than that needed for the free stream temperature to
be reached. This complex relationship might be some of the reason
that controversy has existed on whether heat flux or wall temperature
is the important corrosion variable for some systems.
Unfortunately for the corrosion practitioner, routine laboratory tools for predicting the effect of fluid motion on corrosion in the presence of heat transfer are far less developed than for prediction of the effect of fluid motion on corrosion under isothermal conditions. See Tutorial on CYLEXPERTTM and the Rotating Cylinder Electrode for more information on the latter case. This portion of the tutorial briefly reviews some of the methods that have been reported. Pilot plant sized equipment such as a simulated heat exchanger is mentioned but not emphasized. Heat Flux Apparatus (sometimes with stirring) One of the earliest laboratory methods to be adopted for routine use was a heat flux apparatus sometimes equipped with an agitator (A. O. Fisher and F. L. Whitney, Jr., Corrosion, Vol. 15, p. 257t, 1959 and A. O. Fisher, Corrosion, Vol. 17, p. 215t, 1961). In a later adaptation, the test rig was outfitted for electrochemical measurements (R. J. L. Andon, et al., Br. Corrosion J., Vol. 29, p. 119, 1986). This figure )
shows an illustration of the important parts of the apparatus. Detailed drawings are provided
in the cited literature. This type of apparatus has the following characteristics:
Rotating Disk Electrode The next generation heat transfer apparatus that has been reported is the rotating disk or rotating ring disk electrode with internal heat transfer capability. This device was developed in the 1940’s but applied to corrosion in the 1960’s (P. I. Zarubin, L. A. Poluboyartseva, and V. M. Novakovskii, Zashchita Metallov, Vol. 1, p297, 1965, among others ). It was investigated further in the 1980’s (S. H. Alwash, V. Ashworth, M. Schirkhanzadeh, and G. E. Thompson, Corrosion Science, Vol. 27, p. 383, 1987, among others). It was further developed as a ring-disk electrode. This figure )
shows an illustration of the important parts of the apparatus. The heater has
appeared as a cartridge in the shaft and as a heat transfer medium closer to
the disk surface. More detailed drawings can be found in the above references.
The rotating disk electrode is a step closer to being able to simulate heat transfer effects on corrosion under convective conditions because it, in theory, operates under well-defined hydrodynamics. In fact, data in the first of the above citations showed that heat transfer versus the square root of the rotation velocity was a straight line until higher velocities were reached. Unfortunately, the few studies reported in the 1960’s through the 1980’s do not fully explore the rotating disk electrode for routine corrosion evaluation and simulation under heat transfer conditions. An early paper (P. I. Zarubin, L. A. Poluboyartseva, and V. M. Novakovskii, Zashchita Metallov, Vol. 1, p297, 1965, among others ) attempted to develop equations to relate heat transfer in the rotating disk to that in a pipe. The methodology was to equate the Nusselt numbers in the two geometries by a somewhat circuitous route. The approach may be too simplistic in view of the results found with respect to relating mass transfer between geometries where flow regimes are not the same (e.g. laminar versus turbulent). The rotating disk electrode has not been adopted for routine laboratory evaluation of the effect of heat transfer on corrosion. It may be an alternative for ranking conditions or alloys but validation for such a use is lacking. The ability to add electrochemistry fairly easily could enable additional mechanistic information to be obtained. But, the ability of this apparatus to predict corrosion on a more routine basis may have the limitations outlined below.
Rotating Cylinder Electrode An alternative to the rotating disk electrode is the rotating cylinder electrode. As discussed in the Tutorial on CYLEXPERTTM and the Rotating Cylinder Electrode this apparatus is extremely useful for examining velocity sensitive corrosion under isothermal conditions and single phase flow. Two positive attributes are that the flow regime is turbulent at low rotation rates as is often encountered in the field and the flow field is uniform across the surface. Unfortunately, studies have not been reported of such an apparatus for examination of corrosion under heat transfer conditions though a recent patent application for such a device has appeared (A. Jaralla, "Corrosion Testing Apparatus", US Patent Application 20060162432). The design of the heat transfer portion in that application is derived directly from the rotating disk apparatus discussed above. This figure )
illustrates the components of this type of apparatus.
Application of the rotating cylinder electrode to predicting velocity sensitive corrosion has been discussed in detail in D. C. Silverman, "The Rotating Cylinder Electrode for Examining Velocity-Sensitive Corrosion - A Review", Corrosion, Vol.60, No.11, p. 1003, 2004. 1  (1560k).
As discussed in a subsequent paper (D. C. Silverman, "Conditions for Similarity of Mass-Transfer Coefficients and Fluid Shear Stresses Between the Rotating Cylinder Electrode and Pipe", Corrosion, Vol.61, No.6, p. 515, 2005 1 (403k)),
a procedure might be established to use the rotating cylinder electrode to
estimate the effect of velocity on corrosion in other geometries. The procedure
involves a simultaneous equating of shear stress and mass transfer coefficient in the
two geometries. The question is if this type of analysis can be used with respect to heat
transfer. One of the issues is that the
Prandtl number is not as large as
the Schmidt number .
The ratio of the thermal boundary layer thickness to momentum boundary layerthickness may not
be the same as the mass transfer boundary layer thickness to momentum boundary layer thickness.
Certain assumptions used to relate geometries in the case of mass transfer may
not be valid for all cases involving heat transfer.
The following concept on relating geometries in the presence of fluid flow and heat transfer is hypothesis since corroborating data are not available. The implicit assumption is that the uniformity of flow and thermal boundary layers in the rotating cylinder electrode translates to other geometries also exhibiting similar uniformity. Assuming that a working rotating cylinder electrode with heat transfer capability can be built, the next step is to establish the correlation of the Nusselt number and Reynolds number. in the form 
(25)There are two approaches. The first is to make the measurements of the heat transfer coefficient versus Reynolds number. That approach might be difficult experimentally. The exponents might also be functions of the Prandtl number . The second is to assume the the Reynolds analogy for the relationship between heat and momentum transfer to establish the heat transfer correlation. It is similar to using the Chilton-Colburn modification of the Reynolds number for the mass transfer correlations already developed. These correlations are in the form of equation (25) and are a linearization of a more general correlation. The Reynolds analogy (originally developed for gases in which the Prandtl number is approximately 1) is 
(26a)and 
(26b)where St is the Stanton number as defined in equation (26a), Re is the Reynolds number, Nu is the Nusselt number, Pr is the Prandtl number, and f is the friction factor. The exponent "b" is approximately 0.25 to 0.3 depending on correlation. The exponent c is most likely between about 0.3 and 0.4. Note that the exponent "n" on the friction factor is usually assumed to be 1. That may be an overestimate depending on system. For example, an exponent closer to 0.5 has been suggested for some types of fluids with respect to mass transfer. Once the correlation is developed, how to relate it to pipe flow or other geometries becomes another issue. Such relationships for mass transfer under isothermal conditions have been discussed in the tutorial on CYLEXPERTTM and the rotating cylinder electrodeand in the literature (D. C. Silverman, "The Rotating Cylinder Electrode for Examining Velocity-Sensitive Corrosion - A Review", Corrosion, Vol.60, No.11, p. 1003, 2004. 1 (1560k)).
In that case, equating shear stress and mass transfer coefficient or determining conditions
under which both are equal in the two geometries was concluded to be a reasonable approach
(D. C. Silverman, "Conditions for Similarity of Mass-Transfer Coefficients and Fluid Shear Stresses Between the Rotating Cylinder Electrode and Pipe", Corrosion, Vol.61, No.6, p. 515, 2005 1 (403k)),.
But, there is no reason to believe a priori that equating shear stress and heat transfer coefficients
is appropriate without experimental corroboration.
Though the rotating cylinder electrode is a promising technology, its use for examining corrosion under heat transfer has not been demonstrated and theoretical underpinnings are required to make it a routine tool. Some additional possible shortcomings are as follows.
Heat Exchanger Simulation Reports have surfaced periodically on attempts to simulate corrosion in a heat exchanger in a laboratory scale apparatus. But, to date, no laboratory device has been developed for routine use that can fit multiple situations. Equipment has tended to be "pilot plant" sized. Some of the issues that can impact such development of a laboratory scale apparatus are as follows.
Previous Page: Natural Convection 1 © NACE International publication and year shown in citation above. All rights reserved. Displayed with permission from NACE International, Houston, TX (http://www.nace.org). Published in Corrosion, in the month and year shown in the citation above. |
David C. Silverman, Ph.D. - Primary Consultant |