As a chemical engineer i learnt fluidynamics in the faculty, as a chemist i am making a doctorate in computational chemistry at molecular level. Nowdays , how close are related CFD and molecular computational chemistry ? We can predict molecular propieties with computational chemistry, and simulate heat and momentum transfer with CFD, where are the links between both. Could someone suggest me something to read, contact or something else.

Thanks for your time and best regards.

marc

the link in between chemistry and CFD is the modeling of chemical rates (as a function of local conditions).

Most schemes of chemical kinetics involve several hundreds of elementary reaction, including reaction schemes involving short-living radicals. At the opposite, CFD can only track a few species (bieng convected, diffusing and reacting). The need is for simplified chemical schemes. Typically a 10 reaction mechanism is achievable within CFD computations.

The only practical way nowadays to calculate complex chemistry is to use the local equilibrium hypothesis, when suited, and to build a table relating local state and a progress variable. Several articles on combustion describe such features. But when the transit time trough the flow domain is of the same order of the chemical time, we need to go to full chemistry in the frame of simplified chemical kinetics.

Marc,

They are very interdependant. It is now possible to calculate Arrenhius rates using theortical models. Also, most CFD codes use computational chemistry (or combustion) codes like Chemkin, Stanjan, or NASA's CET 93. There is also a great need to compute transport properties for all sorts of emissions realed compounds. Finally, there still isn't a good understanding of the turbulence / chemistry interaction. All sides are working on this.

Hope this helps,

Thanks, Tony

For information on conenction of complex chemistry with CFD, see http://www.pfd.ie. This describes PFDReaction, a chemical source term calculation program that works with both Fluent and CFX, as an add-on. It calculates the source terms based on pure kinetic information, combined with a multi-scale mixing model. Kinetics all the way from infinitely fast to very slow (relative to mixing) can be accomodated, with up to 50 species and 50 reactions. There are some restrictions, e.g. there muxt be exactly two streams of different composition entering the reactor. Combustion and pollutant formation can be modelled, as well as chemical reactions typical of the chemical and pharmaceutical industry in stirred tanks and static mixers. The next major release will appy to multi-phase reactions, e.g. near gas-liquid (bubble), liquid-liquid (drop) and solid-liquid (particle) interfaces.