Fluid Flow Analysis and Pipe Simulation

How is FloCAD Different from CFD?

Nusselt and Reynolds vs. Navier and Stokes

FloCAD is a 1D thermohydraulic (fluid network modeling, or “FNM”) expansion to the solid/surface thermal modeling tools within Thermal Desktop and RadCAD®. The result is a unique combination of 3D thermal/structural modeling plus 1D fluid flow simulation.

Many engineers are unfamiliar with this approach compared to the many commercial CFD (computational fluid dynamics) packages available. "How is FloCAD different from CFD?” is a common question. Some of the FloCAD capabilities, such as fuel motions within a multi-compartment tank, can even look like a CFD result.

Fast-solving correlation-based methods such as FloCAD have important advantages that will persist even as computers and CFD methods continue to improve in the future. This isn't an "either/or proposition:" FloCAD doesn't compete directly with CFD since both types of fluid modeling have their own strengths for certain applications. In fact, an integrated FloCAD/CFD approach can be very useful for many classes of fluid modeling problems.

Advantages of the FloCAD Approach

FloCAD uses a fast-to-build, fast-to-solve empirical approach: the exact details of the geometry are often avoided by applying heat transfer and pressure drop correlations such that a one dimensional flow field is adequate. Simply put, the FloCAD approach eliminates the need for extra meshing. This choice has many repercussions:

• Cost: The purchase price of the FloCAD suite of tools is much less than that of comparable CFD software. In addition, the time required to learn and retain FloCAD is reduced, as is the time required to build and change models.
• Speed: The solution speed of FloCAD is orders of magnitude faster than that of CFD approaches: meaningful results can be obtained within minutes on a single PC.
  • This means being able to ask “bigger questions” at either higher levels of assembly or what-if and sensitivity studies using parametric modeling and Advanced Design Modules (for optimizing, calibrating to test, statistical design including tolerancing, etc.).
• Phenomena: Transient system-level analyses are not a problem for the FloCAD approach, and it makes quick work of systems that are poorly suited for 3D CFD methods, including ducted air or coolant flows, heat pipes, and two-phase flow.
• Less accuracy? There is a common presumption that CFD methods assume less than correlation-based approaches, and are therefore more accurate. Correlation-based methods like FloCAD do make assumptions such as fully-developed flow profiles, and such default assumptions are not always applicable. Those discrepancies must be overcome by additional guidance from the user. However, it is less commonly recognized that CFD methods struggle with heat transfer solutions because heat transfer is estimated based upon the flow field estimation: errors in heat transfer coefficients on the order of 20% are not uncommon when using CFD. Resolutions of this problem include use of even more fine meshes near the walls (causing even less flexibility and slower solution speeds) and, ironically, application of empiricisms.

Advantages of the CFD approach

The main advantage of CFD over FloCAD thermohydraulic methods is that certain classes of 2D/3D flow problems that are simply not amenable to 1D methods, or at least require too much user involvement (for example, correcting for entrance length effects or specifying losses influencing flow splits) to exploit 1D flow solution methods and available heat transfer correlations.

Nonetheless, many engineers continue to use both methods for such problems. Essentially, they use a CFD solution to develop a correlation for use in codes such as FLUINT: using CFD runs to define a flow split, a pressure or velocity boundary condition, an effective loss or heat transfer coefficient, etc. Using these as inputs to SINDA/FLUINT enables modeling of transients, parametric solutions, optimization, calibrations to test data, two-phase flow solutions, and so forth.

CFD programs can produce postprocessed 3D flowfields and streamlines, which certainly have tremendous appeal in presentations and technical papers, but they can also help build intuition.