CFD (computational fluid dynamics) is an engineering tool for modeling and simulating the behavior and performance of physical systems having thermal phenomena. More precisely, CFD is a numerical method for iteratively solving a system of equations consisting of,
These equations (called in part the Navier-Stokes eqn.s) govern fluid flow and mixing, heat and mass transfer, and the creation and destruction of chemical species. CFD is nearly always implemented on a computer using specialized software written for this purpose.
By creating an engineering model and running a CFD analysis of a thermal system on a computer, an analyst can gain insight into the dynamic behavior of physical system that is often very difficult, time-consuming, and expensive to achieve using experimental methods. Armed with this insight and the ability to quickly conduct parametric studies via computer simulation alone, the designer has the opportunity to efficiently create:
Ideally, only one (1) physical system will require prototyping and subsequent validation testing to prove the desired design and performance goals are met.
When a client contacts us, we have a brief phone discussion to understand and evaluate the nature of the inquiry and get some basic information. Then, if prudent, we arrange for an on-site meeting at the client’s office to gather more information about the application and the client’s needs. If the problem is relatively simple, Computational Results may provide a written quote for the work after this initial meeting. If the project is large or complex, Computational Results may ask for an additional meeting to answer open questions before providing a quote, if appropriate.
This varies greatly depending on the size and difficulty of the project. After an agreement has been struck, sufficient information provided, and initial payment received, the duration of the project will be estimated. Computational Results will attempt to meet reasonable project completion goals requested by the client.
Due to the location of the computers, software, and staff, all work on a project, i.e. the actual modeling, analysis, and report generation, w/ exception of information gathering, will typically take place at, or be coordinated out of, the office of Computational Results.
Computational Results charges a flat fee for work based on the project requirements and goals. Certain additional and unforeseen expenses, e.g. transportation, lodging, and food, related to travel to the client’s site will be billed separately.
Computational Results will use the software tools that we feel are most appropriate to solve the problem. We are not limited to using a particular CFD code.
We have experience with, and access to, a number of:
In addition, we have the knowledge and ability to write and modify CFD codes to simulate unique situations.
CFD is an approximate numerical solution method for complex, real problems. Certain engineering simplifications and mathematical approximations are inherent to the analysis methodology. If the problem is skillfully modeled and the inputs are reasonably correct, the results of the analysis will be valid within the limits of typical engineering accuracy.
Among the largest contributors to the inaccuracy of the solution are the user-supplied (possibly client specified) inputs, i.e., boundary and initial conditions, to the analysis. In actuality, these inputs are often not well-known and are difficult to characterize or determine with any absolute certainty. Thus, they are frequently estimated or assumed.
The solution of a CFD simulation should be thought of as an engineering prediction. For the reasons explained above, all results from a CFD analysis should be regarded as tentative until validated through experimental testing.
Testing may show that the CFD model or input conditions require adjustments to reflect actual performance. After validating the accuracy of the results of an analysis through testing, small changes can be made to the input conditions of the model with fairly good assurance that the results will reliably predict performance. Nevertheless, we always recommend experimental validation of the results of a CFD analysis.

