Analysis? An introduction to the Unsung Hero of Fluid Simulation
w w w . m e n t o r . c o m
What is System Level Thermo-Fluid Analysis? An introduction to the Unsung Hero of Fluid Simulation
Introduction When the history of fluid simulation is written, a justifiably large proportion of the tome will focus upon the development of solutions based on, or derivatives of, the Navier-Stokes equations. In fact, what are commonly recognised as three-dimensional methods will likely dominate to the exclusion of all else. As with many histories, this wouldnt be representative. The reason is simple: there is a type of fluid simulation that adopts a very different approach that has been serving a range of industries very well for over 30 years. It has, and continues to, enabled the safe and efficient design of fluid systems ranging widely in both scale, fluid type and engineering application. This type of simulation is frequently but erroneously termed 1D system CFD. This paper will explain some of the fundamentals of the approach and demonstrate why system-level thermo-fluid analysis is a more accurate description. Its hoped that it will help shed some light on both the philosophy that underpins the system approach and how and why it is used.
A World of Systems, Simulation for that World
In the world of fluid dynamics, the generally accepted view of computer simulation is that of a three-dimensional geometry, immersed in a computational domain, see Figure 1. This domain is discretized, and the thermodynamics of each cell is approximated by the chosen solution method (of which, the Navier-Stokes equations is the most commonly used).
Figure 1 Three-dimensional Simulation of Electronics Housing
w w w. m e nto r. co m
[7 ]
What is System Level Thermo-Fluid Analysis? An introduction to the Unsung Hero of Fluid Simulation
The reason that this is the most widely recognized approach to Computational Fluid Dynamics (CFD) is because it has proven to be an extremely versatile and effective one. However, it is one best suited to isolated geometries with the effects of any surroundings reduced to idealized boundary conditions. Simulation of more realistic systems, for example the effect of slip-streaming an upstream vehicle on a race-cars aerodynamics, has been limited by the available computational power. Taking another example, for engineers concerned with the flow of fluids through pipes, it would be an approach that could supply a great deal of information about the nature of the flow through a conduit, and perhaps even a conduit-bend-conduit system; but what of a more extensive network? Perhaps one that begins with a series of pumps and then goes on via a number of valves, fittings and branches to a number of delivery points several hundreds of kilometres downstream. Not only does this represent a significantly greater computational challenge to the approach described above, the simulation would require a degree of geometrical knowledge of the individual components that may not yet exist. The paradox is that as that knowledge becomes available, the range of useful applications for the resulting simulations diminishes; pump selection via simulation is of little value if details of pump geometry are required up front to run that very simulation! Finally, the effort required by a user to set up such a simulation and the time taken to calculate a result would be prohibitive for any practical application. This class of problem is a broad one. Not only does it cover large scale water, oil or gas distribution networks, it also includes engine lubrication circuits, fuel movement and supply on aircraft of all sizes, ballast systems on ships and submarines and ventilation networks of all types to name but a few. Its almost more difficult to think of any reasonable sized technology that doesnt require some type of fluid network.
Figure 2: A System Model of an Aircraft Fuel Network
The question is how best to provide the designers and operators of such systems with the advantages of computer simulation, but in a practicable and useful manner.
w w w. m e nto r. co m
[7 ]
What is System Level Thermo-Fluid Analysis? An introduction to the Unsung Hero of Fluid Simulation
Building Blocks Stated simply, the point of a system of conduits or pipes is to take a product (or products) from one or more sources and deliver it (or the resulting mixture) to one or more receiving points. The delivery point may of course be the source in the case of a closed circuit. Thus, the most obvious characteristic of a pipe system is that it places distinct bounds on the direction a fluid may travel. If the bulk flow is in a direction other than one end to the other, there must be a failure in the system boundary, i.e. theres a leak! As well as ensuring that all the mass that enters the system leaves at the desired point, it may also be that thermal energy must be conserved: thus insulation and fluid/wall temperatures can also form a critical part of the network simulation. From the perspective of the fluid, at its most basic the system consists of a series of connected components1 which are designed to variously: Store a volume of fluid (e.g. reservoirs or accumulators) Add or remove energy from the fluid (e.g. pumps, compressors, heat exchangers) Direct the kinetic energy of the fluid (e.g. pipes, bends, junctions) Control the flow rate of the fluid (e.g. valves, transitions, orifice plates) If it were possible to find a way to characterise the thermo-fluid effect of these components in a way that could be made generally applicable, system simulation would not have to rely on the discretization approach described above. Happily, this is possible. Although there is a slight complication in that not all fluids are created equal.
Thermodynamics and Fluid Dynamics
Our everyday experience of fluids makes it clear that they possess very different characteristics. Walking through fresh air on the way to work is a very different experience from wading across a swimming pool; getting tomato ketchup out of a bottle is a very different experience than the same operation with water. Even the same fluid can look and behave very differently at different pressures and temperatures. For instance, a boiling kettle contains water in two distinct phases as well as in an intermediate two-phase condition. Effectively accounting for the changes in fluid properties and energy content throughout a system is an important but complex task. Accordingly, some thought has to go in to the level of detail thats appropriate to achieve the correct balance between useful accuracy and solution complexity. For the user, complexity can manifest itself in terms of both excessive and intricate input demands and also solution times that extend into hours, rather than seconds or minutes. Balancing the conflicting requirements is a delicate task and as might be expected, there is no one-size-fits-all solution. For example, air circulating through the ventilation network of an aircrafts cabin may be reasonably modeled using ideal gas assumptions, but the same may not be true of a high pressure multicomponent gas line. One option may be to provide the user with the option of perfect, ideal or real gas models. Two-phase and Non-Newtonian fluids can prove more complex again. While Non-Newtonian fluids can be broadly classified into distinct groups according to the relationship between viscosity and shear stress, the behavior of any individual fluid within a group will depend strongly on its rheology. For example, the aforementioned tomato ketchup is a shear thinning fluid, while custard is a shear thickening one, but the range of fluids within each of these two broad classes is extremely wide and varied. Therefore the accuracy of any simulation will depend strongly upon getting accurate rheological data for the exact fluid under consideration. There are of course many other component types than those listed here. Most notably are those that contain chemical reactions, but these basic categories will serve to illustrate the principles for now.
w w w. m e nto r. co m
[7 ]
What is System Level Thermo-Fluid Analysis? An introduction to the Unsung Hero of Fluid Simulation
The physics of phase-change is another extremely complex field. Pipe orientation, fluid velocity and the rate of heat addition will all affect the behavior of the fluid. Perhaps unsurprisingly given the apparent sensitivity of the physics of phase change to external factors, experimentalists have derived numerous correlations to account for heat transfer to and from two-phase fluids. Which model is most appropriate for a given situation will be something that must be judged by the user.
Pressure Drop: Paying Your Dues
If there is no pressure differential, there is no flow. Analogously, a ball placed on a flat surface wont roll anywhere without an external impetus, usually in the form of your booted foot. A pump or compressor performs much the same task in the world of fluids, though perhaps in a slightly different manner. The mechanism is more akin to tilting the surface the ball is resting on. The gradient of this slope is the pressure difference required to achieve a given flow rate. This pressure difference is effectively a toll that must be paid to the laws of thermodynamics. The fee is paid in energy lost variously to friction, heat and noise throughout the system. Therefore the trick is to design the system such that this fee is kept as low as possible. Pressure losses occur due to two broad effects: the first is frictional pressure drop. Moving a fluid through a component is conceptually no different from pushing a block of any material along a surface; the smoother the surfaces in contact, the faster the pace that can be achieved for a given energy input. Thus a smoother pipe interior makes for lower frictional pressure losses. The second element is a little more complex and while the details vary slightly depending upon the exact geometry in question, the principle is the same for all components. To illustrate the effect of secondary flows, let us consider a simple bend.
Figure 3: Flow Through a 90 Bend. Note Downstream Secondary Flows.
The purpose of a bend is to redirect the flow by a given angle. As the flow is forced on to its new course, the velocity distribution across the width of the pipe changes, becoming more asymmetrical with the flow toward the outside of the bend accelerating relative to that on the inside of the bend. This clearly isnt the normal state of affairs for a straight length of pipe, which in turn means that the flow must expend energy reverting back to its equilibrium state after it exits the pipe (see Figure 3). This can take as much as 50-60 pipe diameters downstream of the bend to achieve, meaning that the influence of a simple bend exists well beyond the physical extent of the component itself. w w w. m e nto r. co m
[7 ]
What is System Level Thermo-Fluid Analysis? An introduction to the Unsung Hero of Fluid Simulation
In conclusion, the pressure drop generated by a particular component is the sum of the frictional loss through that component and the energy expended by the flow to revert back to its equilibrium state downstream of it. If these two phenomena are accurately measured in the laboratory, the pressure drop associated with that component can be captured. This is significant as this value for pressure loss captures inherently three-dimensional effects. If the experiment is extended over a range of flow rates, it is possible to build up a model of that component in two dimensions: pressure drop vs. flow rate. This map can then be used to form the basis of a model of that component which can be used in system simulation. But the news gets better again: converting the ordinates of the map to non-dimensional values (Reynolds number vs. loss coefficient) makes the map applicable to any fluid within that Reynolds number range. The key, of course, is to ensure that the source used for the pressure drop measurements is a consistent and accurate one. However, the more engaged reader may have noted an issue: where components are closely spaced, the downstream pressure recovery described above may be affected. What happens in these cases? The short answer is that these component interaction effects (as they are generally known) must be accounted for either by further experimental measurements (often captured as interaction coefficients), or by three-dimensional CFD. This is in fact a good example of where 3D CFD can be used to effectively support system simulation; in this case by providing an understanding of the likely magnitude of component interaction effects for non-standard combinations. In summary, the overall loss of a system may not simply be the cumulative loss of each individual component, but it is entirely possible to account for the effect of component interaction in system level simulation.
Time: The Fourth Dimension
Having established that system simulation does in fact simulate a three-dimensional reality (it is hoped that it is by now clear that the oft used term 1D simulation isnt an accurate description!) its time to consider the temporal dimension. Flow rates and pressures in systems can change for many reasons; it may be a basic design or operational requirement, or the result of planned equipment shutdowns or even unplanned failures at either a component or process control level. Whatever the root cause, its important to understand the response of the system for a number of reasons. At the most basic level, its important to know that the system is capable of operating across the required range. It may also be important to understand the time taken to achieve a given state, such as when filling or emptying tanks. These may even be safety critical issues. For example, imagine a situation where an aircraft fuel system is unable to move fuel between tanks and to the engine(s) at a rate appropriate to the required demand. Rapid changes in flow rate, often caused by component failure or poorly designed shut down procedures, can even create severe surges in pressure. These surges can result in failures in the pipeline and so pose a significant risk to both personnel and equipment. Accounting for such surges often generically termed water hammer requires that the system simulation tool be able to simulate and track pressure waves throughout the system. These high pressure peaks traverse through the system at the speed of sound for that fluid, reflecting off components such as pumps and closed or partially closed valves. A related phenomenon is that of the vapor cavity. Just as water hammer generates pressure peaks, it also creates pressure troughs. If the pressure drops below the vapor pressure of the fluid, a pocket of vapor can form. This can create two major hazards: if the vapor pressure of the fluid is less than the pressure on the outside of the pipe there is a real risk of pipes being crushed by this differential and imploding. This isnt an unlikely consequence; the vapor pressure of water at 20C is only 2% of that of the atmosphere. Conversely, even if
w w w. m e nto r. co m
[7 ]
What is System Level Thermo-Fluid Analysis? An introduction to the Unsung Hero of Fluid Simulation
the pipe walls can withstand the pressure differential, the subsequent pressure peak generated as a result of the cavity collapsing can be extemely severe. In summary, thermo-fluid simulation that accounts for changes in system parameters with time can be essential for understanding both routine operation and safety critical events.
Who, What, Why, Where and When?
The combination of low computational overhead and broad applicability, combined with the ubiquity of fluid systems, really does mean that thermo-fluid system simulation has a place across virtually all industries. Not only that, the applicability extends to any point of the design and operation life cycle. System simulation has been used to assess and understand space vehicle propulsion, process plant operation, vehicle cooling, fire fighting water mains, gas pipelines, surge protection, lubrication of industrial and small scale transmission systems, aircraft cabin ventilation, submarine ballast tanksits an almost endless list. Perhaps its more interesting to consider when and why it might be used. At the concept design stage, when available system information is by definition sparse, simulation can be used to indicate likely system performance. Operating in a virtual environment also means that evaluation of different or novel approaches can be achieved easily and quickly. As system details emerge and more information becomes available, the system model can be adapted to reflect this. The system model evolves with the project which translates to fewer design iterations and a reduction in the number of physical prototypes required. The benefits can continue to be realised once the system is in service: modifications, adaptations, changes in operating point or even decommissioning scenarios can be run through in the virtual environment.
Summary The aim of this article has been to highlight some of the fundamental aspects of thermo-fluid system simulation. It has shown how a system level approach to system simulation captures three-dimensional loss data in a twodimensional performance map. This method of working captures all the required loss information for a component, without the computational overhead associated with a three-dimensional geometry based approach. This in turn makes it feasible to simulate entire systems in both steady state and transient. The versatility of thermo-fluid system simulation means that it can be applied across a wide range of industries at various points of the design and operation life-cycle.
For the latest product information, call us or visit:
w w w . m e n t o r . c o m
2014 Mentor Graphics Corporation, all rights reserved. This document contains information that is proprietary to Mentor Graphics Corporation and may be duplicated in whole or in part by the original recipient for internal business purposes only, provided that this entire notice appears in all copies. In accepting this document, the recipient agrees to make every reasonable effort to prevent unauthorized use of this information. All trademarks mentioned in this document are the trademarks of their respective owners.
