How Does the ANS Module Work?

ANS is built to allow you to approach the automated sizing process in a way appropriate to your system.

You must specify an Objective - minimizing or maximizing a specific system variable. Most commonly, the ultimate Objective is to minimize monetary cost.

The pipes to be sized must be defined - any given pipe may be included or excluded from the automated sizing process. It may not be desired to automatically size every pipe in the entire system independently. It is common that groups of pipes - such as those in a header - must be of equal size. This is accomplished by placing them into Common Size Groups - sets of pipes that are automatically sized but must all be the same size.

The selection of piping is generally restricted to certain materials, sizes, and types independent of fluid flow concerns. Perhaps certain system piping must be Stainless Steel of any common size, whereas other piping in the system can be regular Steel, but is limited to three on-hand diameters. These are represented with Candidate Sets - selections of pipe sizes and types that are allowed.

If the only goal is to minimize cost, the answer is trivial - select the smallest pipe sizes available. Clearly, this is not practical as systems are not built for the sake of being inexpensive, but are meant to accomplish some goal. These goals, such as minimum flows or pressures, are Design Requirements. Any number of Design Requirements can be specified in a system, controlling the allowed fluid behavior.

There are several different Sizing Methods available. These methods approach the same Objective in different ways, with certain methods being more appropriate for certain systems. Agreement between methods is a strong indication that the ideal sizes have been selected.

Analysis vs. Design

Sizing a system to minimize or maximize a system-wide quantity like monetary cost or pipe weight is a complex process with many considerations. AFT Arrow is foremost an analysis tool, but the ANS module is a design tool. While closely related, analysis and design are different in many ways.

Engineering analysis is the process of using accepted calculation methods to predict the behavior of a given system. Analysis requires the full specification of the system before calculations are made. For example, fundamental fluid equations can be used to determine the flow in a pipe of a known size.

When a new system is being designed, the pipe size is not known. Instead, it might be desired to find the size necessary to meet a minimum flow. The engineer can make an educated guess, perform the analysis, modify the size, and repeat the process. The analytical methods are used iteratively to arrive at a final design.

Even for a simple problem, the iterative Specify Model > Analyze > Investigate Results > Adjust Model method can be very inefficient. For simple systems, it is often possible to solve for a design variable by manipulating the fundamental equations. This would allow the process to be completed in one design iteration.

For complex problems direct solutions are impractical or impossible. Instead, the iterative method can be applied with numerical methods. AFT Arrow already automates the "Specify Model > Analyze" portion of the process. AFT Arrow ANS further simplifies the process by using intelligent searching algorithms to automate the "Investigate Results > Adjust Model" steps.

By providing ANS with information on the requirements of a system, it offers a true design-oriented approach, automating a complex iterative process.