Seeing Noise: Acoustic Optimization of Exhaust Systems Through Simulation

Transcription

1 Seeing Noise: Acoustic Optimization of Exhaust Systems Through Simulation A whitepaper discussing the challenges acoustic engineers face and the use of new approaches in simulation technology to find noise sources and improve acoustic performance. Exa Corporation 2016

2 SEEING NOISE: A NEW APPROACH TO ACOUSTIC OPTIMIZATION OF EXHAUST SYSTEMS THROUGH SIMULATION Executive Summary The sound produced by an automobile s exhaust serves, in many cases, as the signature of the brand a false note sounds as wrong to the customer as their favorite song being sung off-key. The exhaust system is also becoming a regulatory noise issue, closely correlated with performance and fuel economy. The noise emanating from an exhaust system comes from two main sources: 1) acoustic pulses generated by periodic combustion pressures in the cylinders and 2) selfnoise generated within the exhaust system by the flow related to complex geometries and airflow paths. The attenuation of acoustic pulses by the exhaust system can be estimated with 1-D acoustic simulations or simple semi-empirical tools to provide some guidance on the exhaust tuning. But these current simulation solutions are not fully capable of predicting the self-noise produced by the exhaust system, nor able to identify its origin with acceptable accuracy or turnaround times that comply with the vehicle development process PowerFLOW addresses this challenge by accurately predicting the turbulent airflow driven by the most complex and intricate geometry details in the exhaust system while simultaneously capturing and propagating the noise inside and outside the exhaust system. Its proven and acknowledged accuracy for the prediction of aerodynamically induced noise is enabled by key properties of the solver: transient, compressible, very low numerical dissipation, and ability to handle complex, detailed geometry. Exa s PowerACOUSTICS Flow Induced Noise Detection module (FIND ) provides diagnostic technology to pinpoint and quantify noise sources produced by the flow. In addition, engineers can listen to sound generated by the simulated exhaust system long before a prototype is built. This technology makes it possible for engineers to quickly investigate the performance of different exhaust designs and architectures in order to achieve the right sound signature and meet noise and fuel economy standards with fewer design and prototyping iterations.»» COVER IMAGE: Insight on muffler/exhaust noise sources is now possible using Exa solutions. Image shows turbulent flow structures generated by a muffler (using Lambda 2 isosurfaces colored by vorticity) Exa Corporation

3 SEEING NOISE: A NEW APPROACH TO ACOUSTIC OPTIMIZATION OF EXHAUST SYSTEMS THROUGH SIMULATION Challenges of Exhaust System Design Automobile manufacturers recognize that vehicle buyers have expectations of how their vehicle should sound whether they are stopped at a traffic light, accelerating to pass on a freeway, or pulling into their driveway and they are committed to providing an exhaust sound signature that strikes just the right note. A Ferrari, a Bentley, a Maserati, a Harley Davidson or a Ford Mustang can for instance be recognized with one s eyes closed just by listening to their sound. As a result, the sounds that a vehicle generates have become an increasingly important aspect of the development process and can be considered as a real art. At the same time engineers must address tightening government regulations limiting exhaust system and muffler noise. For example, the European Parliament recently passed a law that will require auto manufacturers to reduce noise levels from passenger cars, buses and vans by 4 db and trucks by 3 db. The exhaust system is also playing an increasing role in efforts to reduce automobile fuel consumption. It is generally accepted that for every inch of mercury of back pressure (or 3.4 kpa) approximately 1 or 2 horsepower is lost, resulting in an increase in fuel consumption and carbon monoxide and particulate matter emissions. The periodic high amplitude pressure pulses originating in the combustion chambers are a source of noise contamination both inside and outside the vehicle. Mufflers typically use geometry to create acoustic resonance by creating interferences that reduce the intensity of the noise generated by the engine. The high complexity of the geometry required to accomplish this task also tends to generate airflow disturbances that frequently create noise. The primary objective of muffler design is typically to cancel the pulsating components of the exhaust flow while generating as little noise as possible in the muffler, and at the same time keeping back pressure of the muffler as low as possible. These three main goals frequently conflict with each other, creating a daunting challenge for the exhaust system designer. Limitations of Current Solutions Acoustic simulation solutions that treat the exhaust system as an acoustic filter are commonly used to assess and tune the exhaust noise performance in the early stages of the design process. The quality and characteristics of this filter are usually assessed based on Transmission Loss (TL), which characterizes the ability of the system to transfer acoustic waves at various frequencies. These rapid solutions provide guidance on the ability of the exhaust system to attenuate the engine noise but, because they do not include all of the geometry details or complex temperature and flow interactions, provide little or no information on the noise generated by flow inside the muffler.»» Simulation of the engine muffler with Exa PowerFLOW allows vehicle engineers to visualize and optimize turbulent flow structures within the muffler early in the development process 2015 Exa Corporation 3

4 SEEING NOISE: A NEW APPROACH TO ACOUSTIC OPTIMIZATION OF EXHAUST SYSTEMS THROUGH SIMULATION In some cases, engineers attempt to obtain information on self-induced noise by using 3-D Computational Fluid Dynamics (CFD) simulations that take the full geometry of the muffler into account. The problem in these cases is that mufflers have highly detailed geometry, such as sheet metal with tiny perforations, on the order of 2-3 mm while acoustic waves are much larger, for example, 34 cm wavelength at 1 khz. Another challenge is the difference in time scales between the bulk mass flow through the muffler at 10 to 20 meters per second vs. the noise that propagates at the speed of sound at about meters per second. Finally, there is the need in simulation to resolve the six to eight order of magnitude difference between the 102 to 103 Pascal pressure field induced by the mass flow through the muffler and the acoustic wave amplitude typically on the order of Pascal. With conventional flow simulation software, the fine mesh and short time steps needed to address these enormous differences in magnitude and scale result in very long solution times, typically on the order of weeks or even months. Even when engineers have the luxury of waiting for traditional flow simulation to provide a solution, noise sources can only be identified by correlating transient flow properties with an acoustics simulation. Using a traditional post-processor on the transient data, the flow-induced noise sources are not straightforward to identify and can even be completely masked by the turbulent flow, thus presenting challenges for engineers to determine the origin of problems and to solve them. This complex task is much better suited for researchers than design engineers. Exhaust system suppliers and original equipment manufacturers are normally forced to wait until late in the design process and build prototypes to assess the performance of the actual exhaust systems components. At this point they commonly encounter unpleasant noises generated by the muffler geometry, such as whistles, crackles, etc. The physical tests provide little or no information on what specific geometry is causing the noise. Engineers are then faced late in the design process with the difficult task of re-engineering the exhaust system and then building and testing new prototypes with a tries-and-errors approach. These types of problems have been reported by suppliers and OEMs to take 6 to 12 months to resolve and in some cases new vehicles have been released with these problems because no solution was found Exa Corporation»» With Exa s FIND module, the noise sources are easily visualized and the noise-generating design features are identified. Image above shows the noise sources generated by the muffler (using an isosurface of Acoustic Power, a quantity calculated by the FIND module).»» The FIND module allows the quantification and ranking of the various noise-generating areas of the muffler. Image above shows the location of boxes delimiting areas of interest. Chart below shows the Acoustic Power radiated within the system by each of these areas (the color of each bar corresponds to the box color above).

5 SEEING NOISE: A NEW APPROACH TO ACOUSTIC OPTIMIZATION OF EXHAUST SYSTEMS THROUGH SIMULATION A Breakthrough for Exhaust System Simulations PowerFLOW addresses this problem by simulating the transmission losses, while at the same time accurately predicting the noise-generating turbulent flows driven by the smallest, most intricate geometric details of the exhaust system. It uses a proprietary and efficient Lattice-Boltzmann solver especially suitable for transient flow simulations. PowerFLOW uses proprietary turbulence and near wall models, extensively validated for hundreds of cases, resulting in a very high accuracy of the flow and noise predictions. Transient and compressible simulations of the flow are performed on the fully realistic complex geometries, and the resulting noise sources are computed accurately using an innovative approach based on the dynamics of the flow structures. These simulations can be performed in a time frame that satisfies the automotive development cycle, measured in hours or days. The vehicle geometry including the under and upper bodies can be included in the simulation to evaluate the radiated noise levels and the insertion losses in order to assess the performance in installed conditions. The effect of absorbing materials in the system in presence of the flow can be included, leveraging the Acoustic Porous Medium (APM) Exa patented solution. PowerFLOW can also be used as a pure acoustic solver in order to accurately estimate the transmission losses of the exhaust systems and quickly investigate the performances of various designs and architectures. The FIND module generates clear and meaningful 3-D maps of noise sources that can be visualized and intuitively analyzed using PowerVIZ and PowerREALITY. FIND pinpoints clearly the origin of noise and provides the ranking of the different sources in or around the system. This information simplifies the traditional analysis of the data using, for instance, unsteady flow and band-filtered animations or db maps. A new capability introduced in PowerVIZ 5.3b also enables the automatic creation of clusters of noise sources, giving the user clearer insight on the location and strength of the various noise-generating regions present in the system. Case studies A major exhaust supplier wanted to find out why an exhaust system previously designed for a major OEM emitted a loud whistle when the engine reached a specific rpm. PowerFLOW simulations showed that the whistling noise came from the complex interaction of the flow shedding occurring at the openings of a perforated baffle in the muffler and the acoustic resonance of the neighboring muffler compartments. The geometry of the baffles separating the muffler cavities was redesigned to remove this flow shedding, reducing the noise by about dba. Additional changes were made to the perforated hole geometry to reduce problems observed in slotted holes, bringing an additional 1-2 dba decrease. A vital element in designing high-end automobiles is the engineering of the vehicles signature sound character to impact the proper emotional response of the driver and passengers associated with the brand. Engineers of such a brand did not achieve the right sound in the first prototype of the exhaust system for a new model. The testing results provided little diagnostic information to enable them to improve the design. Engineers performed PowerFLOW simulations to understand how noise was produced in the exhaust system and which geometric structures were responsible for the sound quality. Based on this understanding, they ran a parametric analysis of the dimensions of critical geometrical features in order to determine the effect on the vehicle s sound. They evaluated each virtual prototype of the exhaust system by studying the sound power levels in key frequency ranges and also generated simulated audio signal to obtain a meaningful aural evaluation of the quality. Engineers iterated to a design that delivered the right signature sound while meeting other design requirements in a fraction of the time that would have been required using the traditional build and test method. Conclusion Conventional fluid dynamics software cannot predict and identify the self-noise produced by the exhaust systems within a normal vehicle development cycle. By enabling engineers to accurately assess exhaust system and muffler acoustic performance and efficiently optimize performance before a physical prototype is built, PowerFLOW enables engineers to achieve the right signature sound and meet noise and back pressure targets with a higher level of certainty and with less time and effort. The ability to evaluate a wide range of possible designs without having to build a prototype makes it possible to meet more challenging noise targets and determine the ideal tradeoff between acoustic performance and flow resistance to improve engine operation and fuel economy. Engineers can also run PowerFLOW simulations based on government standards to ensure exhaust system and muffler noise levels comply with these requirements routinely. The ability to accurately and quickly evaluate virtual prototypes in the early stages of the design process makes it possible to get the design right the first time, resulting in a decrease of the development costs and in getting the product to market that meets the need of the consumer. Copyright 2016 Exa Corporation. Exa, PowerFLOW and PowerACOUSTICS are registered trademarks and FIND is a trademark of Exa Corporaiton Exa Corporation 5

6 EXA CORPORATION 55 Network Drive Burlington, MA