CAD Embedded with
CFD Makes for Better
Gas Mixing

Diagram 1

Starting point: Simulation of the original Linnox burner design using medium-pressure air swirl and gas injection showed a good quality mixture.

A major design challenge in many applications is injecting gases so that near-ideal mixing is achieved. Uneven concentrations of air and fuel can substantially increase emissions levels and reduce combustion efficiency. Thorough mixing of gas and air eliminates the hot and cold spots in a flame that are responsible for NO2 emissions.

The traditional practice has been to build a prototype or modify an existing product, test it, and then, based on the results, modify the prototype or product, a process repeated until desirable results are achieved—expensive and time consuming. Now CFD tools can evaluate the performance of a large number of potential alternatives in the early stages of the design process.

With CFD software that is embedded in the CAD system, engineers can use simulation in the design phase to examine more alternatives than would be practical with physical prototyping. Sophisticated automatic control functions make it possible to converge to a solution in almost every application without the need for manual tuning. Perhaps the most important function controls the quality of the mesh to avoid one of the biggest reasons for run divergence. As a result, the skills required to operate CFD software are simply knowledge of the CAD system and the physics of the product, both of which design engineers already possess.

The use of native 3-D data places a premium on the quality of the solid model. For an internal flow model with minimum mesh requirements the solids must form a sealed internal space with no leak paths outside the internal flow field. Minute details of the geometry should be eliminated wherever possible to reduce the number of cells in the CFD model. After the geometry is imported, it should be checked for problems using the “check geometry” feature. Highly skewed cells, caused by holes in a thin solid, can be found by performing a trial mesh generation and corrected by increasing cell density where they occur.

The key factor in selecting the right turbulence model is matching the flow features likely to be present in the application with the models available in your solver. The standard model for turbulence is called k-epsilon, a semi-empirical model based on model transport equations for the turbulent kinetic energy and its dissipation rate. Specialized versions of the k-epsilon model have been developed for specific flow configurations.

Design engineers must verify that their models accurately predict the chemistry and physics of the actual mixing process. One approach is to model the current generation of the product and confirm that the model predicts its performance. Then the designer can modify the model, confident that it will predict the performance of the new design. If it’s too costly to interrupt the operation of the current-generation product, it may make sense to build a small-scale model of the product and compare its performance to a simulation model.

These methods were used with the new-generation Eclipse Linnox burner, designed to substantially reduce the energy consumption of the fans that push air into the natural gas burner, and still provide energy efficiency and emissions control. Engineers needed to streamline the design to remove features that helped achieve high levels of mixing on earlier designs but still maintain the proportion of gas to air at 7.5 percent, +- 0.5 percent, throughout the entire mixture duct. Eclipse designers generated the initial burner designs in Inventor, the 3-D CAD software from Autodesk. They used FloEFD embedded CFD software from the Mentor Graphics Mechanical Analysis Division to simulate them.

The simulation results on the initial model showed the concentration of air and fuel throughout the mixture duct, and highlighted the areas where mixing needed to be improved.

After changing the design, designers reran the simulation to determine the impact of the change, paying particular attention to the species or chemical compound distribution throughout the chamber and the pressure drop. With each major variation, they also performed a series of parametric studies to evaluate the impact of each change.

They gained an understanding of design sensitivities that would have never been possible with physical testing. The simulation results showed that the final design provided a pressure drop of only 300 pascals, a 900 percent reduction from existing burners.

Only then did Eclipse build the first prototype of the new design.

Best practices tuned for the requirements of a particular industry can help design engineers avoid analysis mistakes. By following specific procedures, any engineer can optimize a design at a time when changes can be made at little or no cost.

[Adapted from “The Right Mix,” by Ad Heijmans, for Mechanical Engineering, March 2009.]

With CFD software that is embedded in the CAD system, engineers can use simulation in the design phase to examine more alternatives than would be practical with physical prototyping.


March 2011

by Ad Heijmans