Revista ANSYS

VOLUME V ISSUE 2, 2011
EXC ELLENCE IN EN G I NEE R I NG SI M U L AT IO N
BUILDING ON A GLOBAL REPUTATION
IMPROVE PERFORMANCE
PAGE 4
REDUCE DESIGN TIME

PAGE 10
ENhANCE SySTEMS
PAGE 21
Intel Cluster Ready. Intelligent!

Reduce. Accelerate. Simplify.
The Intel Cluster Ready program can dramatically reduce purchasing complexity, accelerate your deployment timeline and simplify management of your HPC cluster. Intel works with ANSYS and OEMs to provide customers with clusters built for ANSYS workloads. Intel Cluster Ready architecture assures application compatibility and ease of deployment. You can be ready to run. Learn more at www.intel.com/go/cluster
Copyright 2011 Intel Corporation. All rights reserved. Intel and the Intel logo are trademarks of Intel Corporation in the U.S. and other countries.
TABLE OF CONTENTS
Table of Contents
Features
4
Mother of Invention
CONSumEr gOOdS
Simulation helps to deliver a milk warmer to market two years ahead of schedule, saving $54,000 in prototypes. ChEmiCAL PrOCESS
Moving Toward Sustainable Wastewater Treatment

Researchers study the use of fluid dynamics to more efficiently remove environmentally hazardous sludge in chemical plants.
10 AuTOmOTivE
Stepping on the Gas
Circuit and field simulation reduces inductance sensor application engineering from three months to two weeks.
13 OFFShOrE
Deep Thinking
ANSYS structural mechanics helps save years in designing the first steerable conductor for enhanced oil recovery.
16 OFFShOrE / ENErgy
Forecasting Underwater Noise
Simulation soundly predicts hydro-acoustics during offshore pile driving.

10
18 CONSTruCTiON
Building on a Global Reputation
NInsight verifies the architectural design of landmark buildings using engineering simulation.
16
21 TurBOmAChiNEry
Breath of Fresh Air
Cummins uses high-performance computing to deliver better turbochargers to market faster by simulating the entire stage.
SImulAtIoN@Work
24 CONSTruCTiON
18
Fluid dynamics keeps data center server racks cooler using less electricity. 26 hEALThCArE
Better Cooling, Hot Savings
Simulation reduces surgery duration and patient risk in treating infants with skull disorders.
24
Cut to the Bone
26
www.ansys.com
ANSYS Advantage Volume V, Issue 2, 2011
TABLE OF CONTENTS
28 OiL ANd gAS
Multiphase simulation can improve performance of oil and gas separation equipment. 31 TurBOmAChiNEry
Separating the Streams
Erosion rate predictions help designers to improve longevity of rotating equipment that must operate in harsh environments. 34 ThOughT LEAdEr
Predicting Turbomachinery Erosion Rates

28
With simulation and analysis, going green offers many benefits.
Green Design
DepArtmeNtS
37 BEST PrACTiCE Gas turbine supplier leverages ANSYS Workbench to speed complex turbine blade analysis. 40 ANALySiS TOOLS
31
Fast Study
Warming Up to Direct Modeling

Advanced CAE model preparation enhances pellet stove design and optimization.
42 ANALySiS TOOLS
Simulation to Reduce Automotive Emissions
Parametric design optimization of selective catalytic reduction systems helps to meet challenges in emission reduction. 44 ACAdEmiC
40
Virtual Speech
Researchers use virtual tools to model and simulate the human vocal tract and upper airway. 47 TiPS ANd TriCkS
Understanding Actuators
Multiphysics system simulation provides insight into the operations of actuators.
44
About the Cover
The striking architecture of the Florence Railway Station, designed by Sir Norman Foster, has a dramatic roof that allows in ample natural light. Because drainage could be a problem during some rainfall conditions, engineering simulation from ANSYS was employed by consulting firm NInsight to improve the roof design. Page 18.
Courtesy Foster + Partners.
47
ANSYS Advantage 2011 ansys, Inc.
www.ansys.com
For ANSYS, Inc. sales information, call 1.866.267.9724. Email the editorial staff at [email protected]. For address changes contact [email protected].
Executive Editor Fran Hensler Managing Editor Chris Reeves Art Director Dan Hart Editors Erik Ferguson Shane Moeykens Mark Ravenstahl Ad Sales Manager Helen Renshaw Editorial Contributor ANSYS North America Support and Services Editorial Advisor Tom Smithyman Designer Miller Creative Group Circulation Manager Sharon Everts
Neither ANSYS, Inc. nor the senior editor nor miller Creative Group guarantees or warrants accuracy or completeness of the material contained in this publication. ANSYS, ANSYS Workbench, Ansoft Designer, CFX, AutoDYN, FlueNt, GAmBIt, polYFloW, Airpak, DesignSpace, FIDAp, Flotran, Iceboard, Icechip, Icemax, Icepak, FloWizard, FloWlAB, G/turbo, mixSim, Nexxim, Q3D extractor, maxwell, Simplorer, mechanical, professional, Structural, Designmodeler, tGrid, AI*environment, ASAS, AQWA, AutoreaGas, Blademodeler, DesignXplorer, Drop test, eD, engineering knowledge manager, ekm, emag, Fatigue, Icepro, Icewave, mesh morpher, paramesh, tAS, tASStreSS, tASFet, turboGrid, Vista, Vt Accelerator, CADoe, CoolSim, SIwave, ANSYS Spaceclaim Direct modeler, rmxprt, pexprt, HFSS, Full-Wave SpICe, Verifeye, Quickeye, optimetrics, tpA, Ansoftlinks, ephysics, Simulation Driven product Development, Smart engineering Simulation and any and all ANSYS, Inc. brand, product, service and feature names, logos and slogans are registered trademarks or trademarks of ANSYS, Inc. or its subsidiaries located in the united States or other countries. ICem CFD is a trademark licensed by ANSYS, Inc. All other brand, product, service and feature names or trademarks are the property of their respective owners.
ANSYS, Inc. | Southpointe | 275 technology Drive | Canonsburg, pA 15317 | u.S.A.
www.ansys.com
CONSumEr gOOdS
Mother of Invention
Simulation helps to deliver a milk warmer to market two years ahead of schedule, saving $54,000 in prototypes.
By Jim Shaikh, Founder, yoomi, London, U.K.
When my first son, Danial, was born, it was my job to take the expressed milk out of the fridge and heat it up while he cried to be fed. My wife said, You are an engineer. Cant you do something useful? To see what would happen, I assigned my engineering consulting business, Intelligent Fluid Solutions (IFS) Ltd., to come up with an easy way to warm liquids fed to babies. The concept of the yoomi selfwarming baby feeding bottle was thus born. I now work full time with the company yoomi, which has sold thousands of bottle warmers.
The yoomi warmer on left and bottle on right. The warmer fits under the bottle cap and liquid is heated as it flows through the channels of the warmer.
By evaluating many simulation prototypes, they doubled the performance of the initial design concept while building only four physical prototypes. IFS saved $54,000 by avoiding the need for 12 additional physical prototypes, and it delivered the yoomi feed warmer to market two years earlier than otherwise would have been possible. When sodium acetate trihydrate crystals are melted above 80 C and allowed to cool, the aqueous solution becomes subcooled. Clicking a metal disk releases the yoomi is a combined package of baby bottle and trapped crystals, causing the solution to crystallize into warmer. At the touch of a button, it gently warms liquid solid sodium acetate trihydrate and to release heat. In the food to the exact temperature of breast milk, quickly and yoomi feed warmer, the milk or formula runs through safely. the design challenge was to take fresh milk from channels and is heated by the solidification process just 5 C to 34 C within one minute of activation across a full before the baby drinks it. the channels are curved to range of baby drinking speeds. increase the amount of surface area in While engineering a way to heat the contact with the milk. Teat Outlet bottle, IFS engineers looked at a number the key to getting the design right (nipple) of chemical and electrical heaters and was determining the geometry and size of Overmould decided to use sodium acetate the channels where the heat is transferred trihydrate. this component is used to the liquid. Addressing this challenge Screw cap Heat in heating pads and hand warmers requires understanding a complex thermal source because it gives off heat when it system in which sodium acetate trihydrate Bottle changes phase; in addition, it can goes through a phase change. It also be easily recharged by boiling. the requires consideration of how different Plug wall challenge was that the initial channel designs might optimize heat prototype heated the nutritional transfer rates to the fluid. Milk fluids to only 17 C, half the needed the original concept design was temperature. created by conducting very basic the engineering team used fluid numerical analyses based on simple Inlet dynamics software from ANSYS geometries that had been studied in to simulate the multiphase fluid published literature, which had been Prototype 3 design of the yoomi system dynamics of the proposed design. developed for heat transfer in laminar
ANSYS Advantage 2011 ansys, Inc. www.ansys.com
CONSumEr gOOdS
flows. the literature designs were restricted to straight channels and, therefore, provided results that did not match physical testing: When the IFS engineering team designed and rapid prototyped the initial concept design, the milk heated up to only half the required temperature. the engineers could have gone back to the drawing board, changed the design, and built another rapid prototype. But each prototype cost about $4,500 and took a week to build. In addition, the properties of the rapid prototypes did not match the production version of the bottle warmer. It probably would have taken 16 prototypes over four years to develop a marketable product using the prototyping method. IFS engineer Dr. Andrej Horvat used ANSYS CFX fluid dynamics software to model the actual channel geometry. Viewing the simulation results, Horvat was able to
a) Milk volume fraction in prototype 3: a) t = 39.3 s, b) t = 51.9 s
b)
a) Milk volume fraction in prototype 4: a) t = 39.3 s, b) t = 51.6 s With each iteration, channel geometry was adjusted to allow for more consistent liquid flow.
b)
substantially improve heat transfer efficiency by redesigning the channels to increase the fluid path. He adjusted channel width and depth to improve heat transfer. using this approach, he was able to increase the temperature of the milk coming out of the device to 22 C better but still not good enough. Next, Horvat built a model that simulated the phase change of sodium acetate trihydrate to better understand how the heat is generated, transferred through the walls of the channels, and passed to the feeding liquid. this required an accurate model of the solidification process. He used a solidification model of the behavioral characteristics of sodium acetate trihydrate from the literature and calibrated it using the experimental data. Horvat then created user-defined functions and embedded them into the ANSYS CFX software. the fluid dynamics tool provided outstanding performance to obtain convergence using time steps large enough to provide meaningful results for this complex analysis. With this accomplished, the team was able to visualize the complete physics, including how the liquid progresses through the warmer channels, how it pushes air in the opposite direction, and how it picks up heat from the phase change material. the multiphase simulation demonstrated that solidification is not a stable process. As the material goes from fluid to solid, it releases energy, but if the energy is not removed, the reaction stops. As the team reduced the number of channels to improve the efficiency of heat transfer, there was a tendency for the liquid flow rate to vary from channel to channel. this meant the temperature was too cold in some channels. the team applied fluid dynamics to carefully balance the flow between channels. the results showed that air plays an important role in the performance of the yoomi warmer. For example, when the bottle is tipped upside down so the baby can drink, the air rushes out of the nipple and into the heating channels just as the milk is entering these channels. In early designs, simulations showed that air was keeping the feed from flowing into the nipple. Horvat corrected this problem by changing the channel geometry to adjust the air and feed flow ratios so the air could flow in one direction and the milk in the other. It would have taken much longer to understand and correct this problem without the help of CFD simulations. modeling the multiphase behavior of the heating compound substantially improved the agreement between the simulation results and physical test results. this understanding of how the whole system works was critical to developing an optimized design in minimal time, and it would have been impossible without fluid dynamics. using the more accurate simulation results, Horvat was able to
ANSYS Advantage Volume V, Issue 2, 2011 5
www.ansys.com
CONSumEr gOOdS
a)
b)
a)
b)
Temperature in prototype 3: a) t = 39.3 s, b) t = 51.9 s
Temperature in prototype 4: a) t = 39.3 s, b) t = 51.6 s
The design of the feed warmer was iterated to bring the milk at the nipple to an optimal temperature.
further refine the geometry and increase the temperature of the milk to 28 C, a significant improvement but still far from the goal. the team believed they could rapidly improve the design if they were able to evaluate a wider range of geometries more quickly. So Horvat performed a series of design studies using the simulation software, and he used the results to build a simpler analytical model that evaluated designs in less time than full-blown fluid dynamic simulation would take. He applied the analytical model to crunch through thousands of possible geometric alternatives, then picked the best designs and performed a full fluid dynamics analysis on these. using this method, engineers were able to increase the temperature to the target of 34 C.
Fluids simulation from ANSYS saved a huge amount of time and money on this project. If the engineers had followed the conventional approach of build-and-test prototypes, they estimate that they would have spent $54,000 over four years readying the product for market. Instead, fluid dynamics simulations correctly predicted the physics of the product and made it possible to meet design objectives in half the time with only four prototypes. When the yoomi bottle promotion was launched on the web, the initial production run sold out in just four days. production has been ramped up to meet demand, and distribution channels are being expanded to include other u.k. and international retailers. n
a)
b)
c)
d)
Temperature in prototype 3, kplug = 10 W/mK: a) t = 39.3 s, b) t = 60 s, c) t = 120 s, d) t = 600 s Temperature of the warmer over a long period of time
www.ansys.com
ChEmiCAL PrOCESS
Moving Toward Sustainable Wastewater Treatment

researchers study the use of fluid dynamics to more efficiently remove environmentally hazardous sludge in chemical plants.
By Quan Yuan, Cyrus Parks, Jen-Hsiang Kao, Jason Ammerman and Kenneth M. Kromka, The Dow Chemical Company, Freeport, U.S.A.
Chemical process applications produce waste, some of which may contain potentially hazardous materials that must be safely removed before liquid effluent is discharged into the environment. Wastewater treatment (WWt) is closely regulated by stringent government standards and is essential for sustainable chemical processes. Indeed, industry leaders regard sustainability as a social responsibility, preventing pollution and re-using water as one of our most precious resources. the goal in such environmental initiatives is to extract waste materials as efficiently as possible by maximizing throughput rates using the least amount of energy. With this in mind, engineers at Dow Chemical, based in the united States, embarked on a few projects to study the use of fluid dynamics simulation in designing and operating WWt systems in chemical processing plants. this represents a huge step forward in an industry that has traditionally relied on replicating past designs in new plants.
As the Dow r&D team demonstrated in this project, simulation provides important insight into critical WWt operations and enables engineers to readily explore alternate configurations to quickly and easily arrive at an optimized design. the simulation method created is intended to serve as a foundation in developing designs for new chemical processing plants as well as in troubleshooting existing designs and retrofitting older plants with upgraded WWt systems. the method focuses on the critical step in WWt operations in which screened and filtered liquid waste is treated in a large aeration basin essentially a large tank with inlets for receiving raw materials and an outlet for draining off treated effluent. Naturally acclimated bacteria within the basin break down the organic compounds into
Float assembly
Draft tube Rows of floating aerators push liquid downward for even bacterial action throughout mixture. Impeller
Wastewater in
Treated effluent out
Thousands of nozzles in pipes disperse oxygen bubbles for bacterial decomposition of waste
Cutaway diagram shows major parts of the sludge basin for bacterial treatment in chemical processing wastewater.
www.ansys.com
ChEmiCAL PrOCESS
Blue isosurfaces for near-floor liquid velocity of 0.2 feet per second to prevent sludge solids from settling in the basin (top). Similar flow can be achieved using only 13 aerators (bottom).
0.9 Normalized DO Obtained from CFD
0.8
0.7
0.6
0.5
0.4 0.400
0.500 0.600 0.700 0.800 0.900 Normalized DO Obtained from Measurement
1.000
There was good correlation between measurement and fluid dynamics calculations of dissolved oxygen in the basin.
aerators are fairly complex assemblies, as each contains a flotation assembly with an electric motor driving a screw-type impeller that forces liquid down a draft tube. In the face of these complexities, the team had to complete the study quickly and attain results that met acceptable levels of accuracy. moreover, procedures had to be standardized so that engineers subsequently using the approach would not have to re-invent the procedure with each application. to model the effects of multiple pipes and gas sparging nozzles, engineers first simulated the two-phase gasliquid action of a single nozzle to determine the gas distribution and the velocity of liquid from bubbles displacing it to the sides and forcing some of it upward. this small domain was then used as a basis to represent the combined behavior profile for the thousands of nozzles in the entire sludge basin. In representing the floating aerators, the team avoided modeling the complex impeller geometry and instead used supplier data on the aerator pumping capacity and draft tube volume to calculate liquid velocity from the tube for different concentrations of bubbles in the basin. modeling the complex gasliquid flow in the sludge involved first determining bubble size and surface area from experimental measurement or estimates based on the turbulent energy dissipation rate. An interphase mass transfer coefficient (the rate at which oxygen is dissolved in the liquid) was then calculated using the Sherwood correlation. Based on this mass transfer coefficient, fluid dynamics simulation determined total oxygen dissolved along with the resulting biochemical reaction rate in the basin. Calculations for dissolved oxygen and biochemical reaction rate were accomplished through the ANSYS
Inlets
a safe material. At the bottom of one of the basins, lengths of pipe with small sparging nozzles blow out air bubbles that transfer oxygen into the sludge to accelerate decomposition of the waste. the decomposition process in another basin is aided by the stirring action of floating aerators, whose impellers push liquid waste together with oxygen bubbles downward toward the basin floor to discourage settling and promote a more even distribution of bacterial action throughout the mixture. Dow researchers had to overcome formidable challenges in modeling these processes. An extremely wide-ranging dimensional scale had to be accommodated in the model because large sludge basins are typically a few hundred feet long and wide and contain thousands of millimeter-sized gas sparging nozzles. turbulent flow of gas and liquid in WWt processes is highly complex and depends on factors such as the way bubbles are ejected into the liquid, oxygen mass transfer as bubbles dissolve into the liquid, the size and shape of nonspherical bubbles, and the biochemical reactions of oxygen in degrading the waste. the numerous floating
8 ANSYS Advantage 2011 ansys, Inc.
Edge effect
Outlet
Simulation shows that the initial position of four inlets to the sludge basin produce significant areas of insufficient dissolved oxygen, called edge effects (top). Repositioning the inlets reduces these effects significantly (bottom). Red indicates high oxygen demand, while blue indicates low oxygen demand.
www.ansys.com
ChEmiCAL PrOCESS
1 0.8 Normalized DO 0.6 0.4 0.2 0
Dissolved oxygen in the horizontal center plane of the basin increases linearly with the distance from the sludge inlet, so engineers were able to position dissolved oxygen probes around the perimeter of the basin to get accurate measurements.
FlueNt user-defined function (uDF) feature that allows customization of the solution in performing specific calculations through DeFINe macros provided for this purpose. All these variables including gasliquid flow, turbulence, dissolved oxygen and biochemical reaction kinetics were combined into a large-domain model consisting of 10 million computational cells representing the entire sludge basin. Subsequent comparison of the basins dissolved oxygen measurements with fluid dynamics calculations showed good correlation, thus validating the accuracy of the approach. this validation gave engineers confidence in using the model in a wide range of simulations that provided insight into design changes for optimizing the performance of the sludge basin. using the fluid dynamics model, engineers could change a few key parameters to quickly study various alternative configurations in arriving at the most efficient design for meeting performance criteria. many of the simulations use ANSYS FlueNt features to create isosurfaces that define boundaries separating regions of various oxygen levels or liquid velocities, for example. Such displays are helpful in understanding the characteristics of a complex system undergoing multiple behaviors. one area of investigation was the optimal positioning of the floating aerators to minimize areas of low near-floor liquid velocities in which solids are likely to settle and prevent uniform mixing of the sludge. using color-coded isosurface displays of constant near-floor liquid velocities, researchers focused on three settling velocities (0.1, 0.2 and 0.3 feet per second) for particles ranging in size from 500 microns to 5,000 microns in diameter. the team immediately saw areas of low near-floor liquid velocity with the baseline uniform arrangement of 15 aerators
positioned in five rows of three units each. A series of iterations indicated that the most cost-effective arrangement to minimize this slow mixing action was to eliminate two of the aerators in rows closest to the basin outlet. this reduction in the number of aerators from 15 to 13 was counter-intuitive, with simulation providing the direction needed in modifying the design. using this new arrangement of 13 floating aerators, engineers next investigated the position of inlet pipes, with three at the end of the basin opposite the outlet and a fourth on a basin side wall. Simulations showed that sludge entering the basin from the single side-wall inlet pipe produced significant edge effects that is, bands of insufficient dissolved oxygen winding their way throughout the basin. repositioning the inlets based on the simulation results helped to avoid these effects. Based on the three-inlet configuration, simulations also indicated that dissolved oxygen progressed linearly from the inlet side of the basin to the outlet side, with all aerators running at the same speed. Seeing this distribution, engineers concluded that a series of dissolved oxygen probes could be spaced around the perimeter of the basin for WWt operators to most accurately measure dissolved oxygen content in the basin. Such measurements are critical to optimal performance of the sludge basin, allowing the oxygen flow rate to be adjusted according to varying oxygen demands throughout the basin. this selected project shows that WWt performance can be improved significantly through simulation-based optimization of aerator positioning and flow rates, inlet sludge placement, and positioning of dissolved oxygen probes. the study demonstrates that the use of fluid dynamics provides Dow with insight into the design and operation of WWt systems and continues to assist the company in ongoing innovation. n
Reducing the flow rate 50 percent for the bottom aerators resulted in a near-uniform dissolved oxygen concentration throughout the sludge basin. Red indicates 1 and blue indicates 0.
www.ansys.com
AuTOmOTivE
Circuit and field simulation reduces inductance sensor application engineering from three months to two weeks.
By Lingmin Shao, Technical Specialist, and Ryan Elliot, Sensor Team Leader KSR International, Ridgetown, Canada
electronic throttle control works behind the scenes to provide a number of benefits in automotive applications. It facilitates the integration of features such as stability control which stops the vehicle from skidding automatic braking systems, cruise control, traction control and pre-crash systems that require the throttle to be moved independently of the position of the accelerator pedal. At the same time, electronic throttle control lowers emissions and improves fuel economy. the technology removes the mechanical link between the accelerator and throttle and instead controls the throttle with an electric motor. An electronic control unit (eCu) determines the correct throttle position based on data collected from two
sensors that track the position of the gas pedal. the motor that controls the throttle is then driven to the required position via closed-loop control from the eCu. kSr International is a leading supplier of inductance sensors used to determine gas pedal position, which allows the electric motor to control the throttle. most inductance sensor applications can be addressed by a common design; however, as kSr expands the technology to different applications with different sensing ranges and packages, a custom design is required. Initially, kSr used trial-and-error methods, which took approximately three months to develop custom inductance sensor designs. more recently, the company has been using Nexxim circuit simulation and HFSS
Aluminum rotor
ASIC: Driving the oscillator signal-processing calibration
Coils: Transmitting coil Receiving coil Reference coil
(CR) (RM) (AM)
Induction sensor architecture
3-D full-wave finite element electromagnetic simulation software from ANSYS to evaluate the complete operation of its inductance sensors. Because evaluating software prototypes is faster than building and testing hardware prototypes, kSr has reduced the time required to engineer a custom inductance sensor application to only two weeks. A critical requirement of electronic throttle control is that a sensor accurately and reliably determine gas pedal position as the driver moves it with his or her foot. Hall-effect sensors provide one option; however, they rely on a permanent magnet whose magnetism might be reduced over its lifetime which might cause the sensor to produce inaccurate readings. Inductance sensors, on the other hand, send alternating current through fixed transmission coils. these devices produce an electromagnetic field that generates eddy currents in a metal rotor connected to the gas pedal or other device. the eddy currents produce an alternating current in receiver coils. the magnitude and phase of this alternating current depend on the position of the rotor. A single application-specific integrated circuit (ASIC) excites the transmission coil and interprets the signal from the receiving coil to determine gas pedal position. one concern with conventional inductance sensors is that inevitable manufacturing tolerances, such as
www.ansys.com
10
Background image iStockphoto.com/Okea
Stepping on the Gas
Electronic throttle control pedal position sensor
AuTOmOTivE
variations in the air gap between the rotor and transmitter and receiver coils, affect the sensors transfer function, which, in turn, may generate inaccurate readings. kSrs innovative design overcomes this problem by incorporating a reference coil designed so that the mutual inductance between the transmitter coil and the reference coil is affected by the air gap to the same degree. then, when processing the signal from the receiving coil, the mutual inductance between the transmitter and receiver coils is divided by the mutual inductance between the transmitter and reference coils. this provides a rotational angle measurement that is independent of the air gap. kSr faces a significant design challenge in designing custom sensor applications that vary significantly from the baseline. these applications typically have a different range of motion, which can require new coils and an application-specific integrated circuit to drive the transmitter coil,
Raw Signal @ Angle = -45deg 2.60 2.55 Y1 [V] 2.50 2.45
Circuit simulation setup
measure the voltage generated in the receiver coil, and calculate the position of the gas pedal or other device whose rotational position is being measured. In many cases, inductance sensor designers must address interference from other components in the customers product. For example, nearby metal parts might affect the magnetic field generated by the transmitter coil and rotor. In the past, kSr engineers developed an initial concept design based on their experience and intuition
and then built a prototype. In some cases, performance fell short of re q u i re m e n t s . p h y s i c a l t e s t i n g provides only a very limited number of data points, so it was often difficult to diagnose the root cause. As a result, the process called for a considerable number of designbuildtest cycles to meet customer requirements and reach an optimized design. the timeconsuming approach was dependent on the engineers skill and dedication. more recently, kSrs use of HFSS and Nexxim software has reduced the need for physical prototyping. the
Output for different angle
5.00
4.00
2.36
197.80
198.00
198.20
198.40 198.60 Time [us]
198.80
199.00
199.20
V(pps) [V]
2.40
3.00
Raw Signal @ Angle = 0deg 2.60 2.55 2.50 Y1 [V] 2.45 2.40 2.36 198.32 198.50 198.75 199.00 199.25 Time [us] 199.50 199.75 200.00
2.00
1.00
0.00 0.00
25.00
50.00
75.00
100.00 Time [us]
125.00
150.00
175.00
200.00
Output vs. Angle 5 4 3 2 1 0
Raw Signal @ Angle = 45deg 2.60 2.55 Y1 [V] 2.50 2.45 2.40 2.36 198.60 198.80 199.00 199.20 199.40 Time [us] 199.60 199.80 200.00 -45
45
Circuit simulation output shows the transfer function of a proposed design.
www.ansys.com
11
AuTOmOTivE
process begins by importing initial design geometry into HFSS from a computer-aided design (CAD) file. the engineer defines the electrical material properties such as permittivity and dielectric loss tangent, permeability and magnetic loss tangent, bulk electrical conductivity, and magnetic saturation along with any boundary conditions necessary to specify field behavior on surfaces. HFSS computes the full electromagnetic field pattern inside the structure, calculating all modes and ports simultaneously for the 3-D field solution. It then computes the generalized S-matrix, reducing the full 3-D electromagnetic behavior of the structure to a set of high-frequency circuit parameters. the HFSS model is inserted as a block into a Nexxim full-circuit simulation, which combines the accuracy of electromagnetics with the simulation speed of circuit simulation. Nexxim provides transistor-level accuracy required for simulating sensitivity analog and wireless frontend circuits as well as robustness and capacity needed to address the diversity and complexity of modern mixed-signal integrated circuit designs. Inductive sensor performance depends on a good oscillator. Fullcircuit simulation makes it easy to determine if the oscillator has a clean wave form and how much current it consumes. Circuit simulation output provides other important design information, such as the signal at various rotor angles. Full ASIC
models can be imported directly into Nexxim in protected or unprotected versions. parametric simulations are often performed to characterize the model in terms of physical dimensions, material properties or various states. the HFSS model is parameterized and characterized over user-defined ranges in parameter space and frequency. During circuit simulation, a multidimensional interpolation is applied to obtain S-parameter data required by the circuit simulator for particular parametric instances of the model. For example, kSr engineers typically perform a parametric study of the air gap between the rotor and transmitter and receiver coils to ensure that inevitable manufacturing variations will not affect the sensor accuracy. Different air gaps require different tank current levels to maintain the same raw signal strength. the output should be independent of air gaps because of the reference coils role (described earlier). the full-circuit model helps kSr engineers to address inevitable issues that arise in real-world applications. For example, steel brackets located near the sensor act as a secondary eddy plate and can have a significant impact on output. In one application, adding a steel bracket, shaft and bolt reduced output voltage by 11.6 percent, an unacceptably high amount. kSr engineers created a series of HFSS models that isolated each of the components. this showed that the bracket alone reduced the
Original coil design (left) and improved design (right)
Steel bracket
Steel shaft
Steel bolt
Adding a hole to the bracket reduced its effect on output voltage.
output by 10.2 percent, the shaft by only 1.3 percent, and the bolt by 0.1 percent. kSr engineers focused their attention on the bracket and attempted several design changes. the team discovered that adding a 20 mm hole to the bracket reduced its impact to an increase of 0.7 percent of output. When added to the effects of the bolt and shaft, the net result was a negligible impact on output voltages, which can be corrected for with standard final programming adjustments. In another application, parametric analysis revealed that changing the temperature of the inductance sensor caused the output to drift. kSr engineers analyzed the root cause by separately analyzing changes in coil temperature and ASIC. this study determined that the coil was the major contributor to sensor temperature drift. the team developed a new coil that reduced the output variation from 2 percent to 0.5 percent. Simulation has enabled kSr engineers to optimize inductance sensor designs prior to building a prototype. the new approach substantially reduces the time required to engineer its products to fit customer applications. the net result is a substantial decrease in engineering costs and a reduction in time to market. n
www.ansys.com
12
OFFShOrE
Deep Thinking
ANSYS structural mechanics helps save years in designing the first steerable conductor for enhanced oil recovery.
By Rae Younger, Managing Director, Cognity Limited, Aberdeen, Scotland
one of the biggest challenges in offshore drilling is accurate placement of the conductor casing. this component is a several-hundred-meter-long tube that is pile-driven into the ground prior to drilling to prevent mud from collapsing around the hole. At offshore locations, soils tend to be relatively soft with highly variable seabed properties; these factors contribute to accurate placement, since traditional conductors follow the path of least resistance. engineering consulting firm Cognity limited has addressed this problem by developing a steerable conductor that can provide real-time accurate positioning. this device must withstand compressive forces of up to 600 tons as the conductor is pounded into the ground; it also must provide an unobstructed bore once it is driven to depth. Soils increase in strength with depth, which increases the moment and loads on the conductor as it is driven into the seabed. By using ANSYS mechanical software in the ANSYS Workbench platform, Cognity engineers doubled the load-carrying capacity of the steering mechanism, allowing the conductor to be maneuvered in very deep soils. In addition, the team finalized the design in five months, a time frame months or possibly years less than would have been required using traditional design methods.
In drilling, each conductor must be positioned accurately to help maximize field production. For example, conductors might be spaced along a 2.5 meter grid at the platform with the goal of driving them into the seabed at an angle, spreading out to cover a predefined area. Since the drilling process weakens the soil, new conductors are naturally drawn toward existing wells which might result in abandoning the conductor if it veers too oorly close to a live well. poorly positioned conductors, known as junked slots, can result in a production company incurring lost time and additional expense in sidetracking them. A worst-case scenario can occur if a conductor is placed so close to an existing well that the
Rendering of steerable conductor
www.ansys.com
13
image iStockphoto.com/sculpies
OFFShOrE
milling tool removing the shoe the blunt nose of the conductor p u n c t u re s a n e a r b y p ro d u c i n g well. Such a scenario may risk an uncontrolled release of hydrocarbons. on behalf of a client, Cognity developed a fully steerable conductor capable of accurate placement in highly variable soil conditions. over the past decade, the industry has trialed designs that passively vary the angle of the shoe in response to changes in soil conditions. But Cognitys design is the first to allow the conductor to be steered in real time from the drilling platform, which enables very accurate control of the final position. the benefits of such a system include possible increased production and re d u c e d d r i l l i n g c o s t s t h ro u g h elimination of junk slots. Design of the new steerable conductor presented major challenges: the most noteworthy is that the device must withstand the enormous forces required to drive a blunt object hundreds of meters into the soil. A traditional design approach would have required numerous full-scale prototypes, each tested to failure a very expensive, time-consuming process. It would have taken several years for the Cognity team to develop a workable design; engineers would have had to settle for the first design that Nonlinear springs were used to represent soil forces met minimum requirements acting on the conductor. rather than aiming to optimize the design. Cognity took a different approach by using ANSYS mechanical simulation software, developing virtual prototypes to evaluate alternative design performance. Cognity selected the ANSYS Workbench platform because of its ability to move new design ideas from computer-aided design (CAD) into simulation, then send proposed design improvements back to CAD critical to meeting the projects tight time schedule. ANSYS Workbench offers bidirectional connectivity with popular CAD systems, including Autodesk Inventor, which Cognity uses. echanical ANSYS mechanical software is also more applicable to design and optimization than other finite element (Fe)
analysis packages that Cognity evaluated. For example, an engineer can set up contacts with a click of a mouse, and these contacts will automatically update when the geometry changes. this feature saved Cognity considerable time in developing the device, which involves large assemblies of moving parts with multiple contact faces. the ANSYS structural mechanics software also provided excellent scalability on nonparallel machines, which helped to support fast turnaround times required for development. one factor critical to success was accurate modeling of the soil. Cognity engineers modeled various conductor concept designs and evaluated their performance when driven into a virtual environment: soil of varying properties. Soil has a highly nonlinear response, providing only compressive resistance under lateral loads. Friction acts on the outer surface of the conductor, creating drag forces that resist axial movement. Soil shear strengths vary with depth and specific location, and Cognity used actual soil test data to increase simulation accuracy. the engineers modeled the soil by using nonlinear springs connected to the conductor, tuned to provide the same stiffness as the soil at a particular depth. mimicking soil, the nonlinear spring provides resistance proportional to the force up to its shear point; from that point on, the force is constant. one of the first tasks required was optimization of the conductors shoe length. During drilling, the operator steers the conductor by changing the angle of the shoe. the shoe moves plus and minus 3 degrees in both x and y axes. A longer shoe better provides maneuverability in soft soil; however, it increases both the reaction force and resulting moment on steering components that connect the shoe to the rest of the conductor. Cognity engineers modeled the conductor being driven into the ground with a 600-ton force from the hammer, then used FE analysis results show stresses analysis results to establish on the tendon.
14
www.ansys.com
OFFShOrE
Stresses on radial locking pads that hold HDH in place
the maximum generated moment and loads from the soil reactions at the shoe. this helped Cognity engineers to identify the loads on the critical steering assembly. the next step was to apply these loads to the conductors principle components so they could be optimized to resist the forces. one critical component is the hydraulic deflection housing (HDH), a 4-ton assembly within the 27-inch bore of the conductor. the HDH is responsible for holding the shoe in position and resists the forces generated by the soil. Analysis showed that the shock loading on this assembly is of the order of 150 g, which necessitated a 600-ton-capacity locking mechanism to hold the HDH in place. After the conductor is driven into the ground, the HDH is recovered, inspected and refurbished so it can be used again.
Cognity applied ANSYS mechanical software to determine the stresses and deflections on the forging that makes up the HDHs body. the primary measure of its performance is its moment capacity, which identifies the ability to generate side load at an equivalent length. engineers optimized the shape of the HDH, increasing its stiffness by adding material to high-stress areas and removing material from low-stress areas through an iterative process. the HDH protrudes into the shoe; it is tapered to provide clearance for the shoe to move in both the x and y axes. Guided by structural mechanics analysis results, Cognity engineers found a more efficient way to taper the HDH and added supports in high-stress areas. As a result, the team was able to double the length at which the HDH connects to the shoe, effectively doubling the systems load-resisting capacity. the original design used custom hydraulic cylinders that cost about $160,000 each and required four months for delivery. using engineering simulation, Cognity engineers demonstrated that the custom cylinders could be replaced with the internal parts from off-the-shelf hydraulics that cost only $7,000 each and could be delivered within one month. For the overall project, Cognity was able to complete the design in only five months, approximately 70 percent less time than would have been required using conventional methods. n
HPC Expedites the Design Process

the use of high-performance computing was critical to meeting delivery-time requirements of this project. Cognity runs structural mechanics software from ANSYS on a Dell t7500 workstation with 12 cores and 24 GB rAm with rAID 0 SCSI drives for optimal disk speed. A typical model with about 750 k elements and many contacts can be solved in an hour or less, compared to about six hours without parallel processing. parallel processing makes it possible to evaluate five to 10 design iterations per day, enabling Cognity to rapidly improve their design.
Stress analysis of the HDH helped Cognity engineers double system capacity by optimizing design.
www.ansys.com
15
OFFShOrE / ENErgy
Forecasting Underwater Noise

Simulation soundly predicts hydro-acoustics during offshore pile driving.
By Ulrich Steinhagen, R&D Project Coordinator, MENCK GmbH, Kaltenkirchen, Germany Marold Moosrainer, Head of Consulting, CADFEM GmbH, Grafing, Germany
MENCK hydraulic hammer
Across the vastness of our planets oceans, localized sounds originate from many different sources, both natural and man-made. the man-made sources for example, from ship traffic, drilling, mining and sonar equipment have significantly added to underwater noise in recent decades. In the growing offshore wind power industry, sound emission related to installation and
Ram
Anvil
Follower
Pile
Closeup of hammers meshed region including lower section of ram, anvil and top of the follower
Water
FSI
Two-D axisymmetric simulation model in ANSYS Mechanical software. Components of the hammer and pile system along with the water and seabed zones that they inhabit; zone of interest for FSI is indicated with a line.
operation, especially underwater, is a growing concern due to its potential impact on nearby aquatic life. In most cases, foundations for massive offshore structures such as wind turbines are formed by driving piles into the seabed with hydraulic hammers. the German company meNCk GmbH has a long history of developing, manufacturing and operating such hammers in water depths up to 2,000 meters (1.25 miles). predicting hydro-acoustics during offshore pile driving is, therefore, of great interest to installation contractors who must comply with tight sound-emission thresholds. For example, the regulatory limit in Germany for underwater sound exposure level at a distance of 750 meters from a construction site is 160 decibels (dB) at a reference pressure of 1 micropascal (1x10-6 pa) [1]. knowing the sound emission prior to construction helps contractors to select and design noise protection systems such as air bubble curtains or air-filled cofferdams around the pile that will meet local project requirements [2]. reducing underwater noise, however, remains an ongoing subject of research, as no single system is appropriate for all situations. With this background, a meNCk research team initiated an application project with computer-aided engineering software and services company CADFem to use ANSYS simulation tools to numerically predict underwater sound emission. transient structural analysis of the driving impact is commonly performed to evaluate mechanical characteristics of highly loaded hammer components, such as the ram, anvil, adapter plates and followers. meNCk engineers expanded this original simulation to consider noise propagation by two-way coupling of the pile vibrations and water pressure using the acoustic elements in ANSYS mechanical software.
www.ansys.com
16
OFFShOrE / ENErgy
the two-dimensional axisymmetric simulation consisted of the ram, anvil, follower, pile, soil and water. Coupling the water domain to the structural elements of the pile was made possible via ANSYS mechanical fluidstructure interaction (FSI). the meNCk team set the rams initial velocity to achieve the desired impact energy, with all other components being at rest. Boundary conditions included acoustical reflection at the seabed and acoustical absorption at the outer boundary, while the elastic soil properties were modeled as lateral springs. researchers applied a simple absorption boundary condition because, when only the short-term near-field sound propagation is of interest, reflection is not a crucial issue for transient analysis of acoustics problems. the team additionally approximated the waterair interface as a free surface with zero pressure, which is appropriate for a non-rigid boundary. meNCks engineers set up the simulation in the ANSYS Workbench environment using ANSYS parametric Design language (ApDl) to control FSI and acoustics parameters. FluID29 acoustic elements allowed modeling the sound field to cover modal, harmonic and transient solutions. the theory of acoustic waves underlying the FluID29 element approach is based on the same fundamental equations as computational fluid dynamics (CFD): conservation of mass and momentum. However, assumptions such as zero-flow velocity and inviscid, compressible fluid properties were made that result in a linearized acoustic wave equation. A linear equation is reasonable because, even for very high sound pressure levels, the acoustic pressure variations generally represent only about 0.2 percent of ambient pressure. In concert with the FluID29 elements, additional displacement degrees of freedom supported the interface to the structural domain. In this way, meNCk could model the full coupling between the fluid (acoustic) and structure
Comparison of Measurement and Calculation at 245 m from Pile 8000 6000 4000 Sound Pressure [Pa] 2000 0 -2000 -4000 -6000 -8000 0.010 0.015 0.020 0.025 0.030 Time [s] 0.035 0.040 0.045 0.050 Measurement Calculation
Comparison of measured and calculated underwater sound pressure at a distance of 245 meters from the pile. Knowing the sound propagation law for this region, the sound pressure at 750 meters can be calculated and converted into decibels (dB).
domains to account for the sound radiated by a vibrating structure and, at the same time, consider the additional load of this sound pressure field onto the structure [3]. the near-field solution in the vicinity of the pile could then be used to predict the sound pressure level in the far field by means of an additional model that accounts for the effects of prevailing ocean characteristics on sound propagation. For this additional model, the meNCk group used analytical relations based on test data that were available for sound propagation as a function of distance accounting for water depth and seabed properties. Numerical results from the simulation have been validated for the installation of the monopile on the FINo3, a government-sponsored wind energy research platform in the North Sea [4]. the comparison of measured and calculated sound pressure at a distance of 245 meters from the pile showed good correlation of the first pressure peaks amplitude. Beyond this initial analysis work, however, further validation is required: the peak sound pressure level observed near the pile is relatively high compared to the ambient underwater pressure, which might violate linear wave theory. A full FSI analysis to couple ANSYS mechanical with ANSYS CFD fluid flow simulation software without the stated typical assumptions of linear acoustics may be applied for this purpose. n
References
[1] Abromeit, C. Licensing Requirements and Conditions in the German EEZ, Federal Maritime and Hydrographic Agency (BSH). Proceedings of 24th Conference of the European Cetacean Society, Stralsund, Germany, March 2010. [2] Nehls, G.; Betke, K.; Eckelmann, S.; Ros, M. Assessment and Costs of Potential Engineering Solutions for the Mitigation of the Impacts of Underwater Noise Arising from the Construction of Offshore Windfarms; BioConsult SH report on behalf of COWRIE Ltd.: Husum, Germany, September 2007. [3] Moosrainer, M. Analyzing Vibration with AcousticStructural Coupling. ANSYS Advantage, 2009, 3(2), pp. 4042. [4] Steinhagen, U. Transient FE Simulation of the Hydro-Acoustics during Pile Driving. Proceedings of ANSYS Conference & 27th CADFEM Users Meeting, Leipzig, Germany, November 2009.
Underwater sound generation and propagation shown as a sequence of snapshots in time. Within a steel pile, the speed of sound is about 5,000 meters per second, while the speed of sound in water is about 1,500 meters per second resulting in radiation patterns and specific inclination angle.
www.ansys.com
CONSTruCTiON
Building on a Global Reputation

NInsight verifies the architectural design of landmark buildings using engineering simulation.
By Michael Stadler, Research Scientist, NInsight, Graz, Austria
In the world of architecture, practitioners like Frank Gehry are pushing the limits of building design with their use of novel shapes and innovative materials, while engineers face the practical task of bringing these elaborate visions to life. Architectural and structural engineers not only ensure the integrity of the architects original concept, they also address such pragmatic issues as structural soundness, occupant and pedestrian safety, wind and rain loads, and the long-term effects of weather on building materials. the Austrian firm NInsight has developed a worldwide reputation for helping architectural engineers conduct research, analyses and simulations that verify building designs before significant funds are invested in construction. Founded in 1992, the organization studies and modifies architectural designs and building materials, among other projects, in a virtual environment with the aid of software from ANSYS. one of NInsights most visible projects is the Guggenheim museum in Bilbao, Spain, designed by Gehry and immediately lauded as an architectural landmark when it opened in 1997. the Guggenheim boasts one of the worlds most recognizable designs, with its titanium surface and organic contours. But when Gehry conceived his architectural vision in the mid-1990s, there were many unknown factors, including how the unique shape would affect wind flows and velocities around the structure, whether the single-support canopy at the entrance would remain stable over time, and how the new titanium alloy would withstand the coastal environment.
Austrian engineering research firm NInsight used ANSYS software to verify some critical design elements for the world-renowned Guggenheim Museum in Bilbao, Spain.
NInsight modeled the Guggenheim Museums facade with spline surfaces, which helped to create a high-quality mesh for subsequent simulations. This hybrid tetrahedra/hexahedra mesh with prisms in the inflation layer consisted of approximately 3,600,000 discrete elements, leading to an extremely accurate and realistic simulation environment. This simulation shows the streamlines and pressure distribution for typical wind conditions in the city.
ANSYS software was absolutely critical to the engineering verification performed for the Guggenheim Museum Bilbao because of its ability to handle complex shapes and free-form surfaces.
18
www.ansys.com
CONSTruCTiON
using ANSYS ICem CFD and ANSYS CFX software, NInsight simulated wind patterns, streamlines and pressures around the structure and suggested small design modificationssuch as slight changes in the shape of the building near ground level and modifications to the entrance canopy. the original shape of the canopy led to undesirable effects, including wind velocities that were too high at the pedestrian level. engineers addressed this issue by modifying the shape of the canopy so that pedestrians could move safely and comfortably around the buildings perimeter. ANSYS ICem CFD software enabled NInsight researchers to create an incredibly detailed geometric mesh of the Guggenheim structure. the meshing process was conveniently driven by automatic curvature, proximity and quality criteria. this helped to avoid extensively distorted or skewed elements that would degrade the quality of the simulation, lead to slow convergence of results, or make it impossible to attain results at all. the powerful capabilities of ANSYS CFX empowered the NInsight engineering team to carry out wind flow simulations with a flow solver for discrete NavierStokes equations using the k- model for turbulence. Architectural researchers also simulated the long-term effects of weather on the exteriors titanium alloy, which resulted in slight changes in the materials composition that would improve its resistance to erosion. to predict erosion effects, they used the mathematical model by Finnie, which relates wear to the rate of kinetic energy of particles impacting the surface. For particleair momentum coupling, NInsight used the model by SchillerNaumann found in ANSYS CFX software. ANSYS technology was critical to the engineering verification performed for the Guggenheim because of its ability to handle complex shapes and free-form surfaces.
The visual appearance of the Guggenheim facade particularly the reflective properties was critical to the ultimate success of the building. Because Bilbao is in a coastal region, it was necessary for NInsight to analyze the long-term erosion effects of sand particles on the facade. In the simulation shown here, the red highlights indicate regions where the effects of erosion would be most pronounced. Based on this analysis, engineers studied several different material compositions to find the optimal resistance properties.
Without the software, NInsights analysis would have entailed creating multiple physical models for testing under changing conditions in wind tunnels and other physical environments. the simulations would have been virtually impossible to re-create with physical prototypes and tests, due to both the costs and time involved. Instead, the architectural team quickly and intuitively created more than 150 virtual test cases that subjected the Guggenheim design to a variety of changing conditions and scenarios which would have taken years to duplicate with real-world testing. the software saved the company time and money and created a much greater degree of confidence in the safety and performance facets of this landmark structure. the parametric studies originally were automated with custom-designed scripts,
ANSYS and NInsight: a Longtime Collaboration

NInsight has leveraged landmark projects to build its global reputation for supporting the research needs of engineering firms and ANSYS software has always been the firms tool of choice. the algorithms of ANSYS software are the most stable I have ever seen, which has allowed our research team to push the limits of engineering simulation just as our architectural clients are challenging the boundaries of building design, said Stadler. He noted that ANSYS software makes it possible for NInsight researchers to model complex shapes and free-form surfaces, like the Guggenheim, as well as to perform multiphase simulations required by the Florence, Italy, railway station project. In 18 years of working with this software, we have never encountered an engineering simulation challenge that ANSYS tools were not equipped to address, said Stadler. While gaining a reputation for simulation of architectural projects, NInsight also simulates the performance of biomedical devices, engines, and a variety of industrial and consumer products. Whether we are modeling an entire building or a small coronary stent, ANSYS tools help our clients to see the actual effects of complex systems and mechanisms that would otherwise be difficult, time consuming and prohibitively expensive to measure, said Stadler. engineering simulation helps our customers to understand exactly what is going on, both quickly and cost effectively. this knowledge is indispensable for the successful design of a range of systems, from architectural landmarks to medical devices, consumer products and industrial components.
www.ansys.com
19
CONSTruCTiON
because the ANSYS Workbench platform was not yet available when the building was analyzed. However, with the introduction of ANSYS Workbench, these studies became even easier to manage. With bidirectional CAD connectivity, automated meshing, a project-level update mechanism, pervasive parameter management and integrated optimization tools, ANSYS CFD delivered unprecedented productivity in analyzing multiple design variations.
Extraordinary Designs ... Practical Considerations more recently, NInsights architectural research team worked to verify some design elements for yet another landmark structure: the breathtaking new Florence railway Station, commissioned by the Italian government and conceived by world-renowned architect Sir Norman Foster. Scheduled to open in 2016, this innovative structure is already being hailed for unique features such as its dramatic arching roof, which allows an abundance of natural light into the interior spaces of the building. the roof is the stations most striking feature, and it is the most problematic from an engineering perspective. the roofs shape necessitates the collection and safe distribution of massive amounts of rainwater. NInsight was commissioned to analyze the roofs open-channel drainage system and to perform a series of complex rain flow simulations. these analyses were aimed at studying the roofs performance under different rainfall and drainage situations, the erosion patterns of its many glass surfaces, and the effects of falling rainwater on pedestrian safety. Analyzing the roofs drainage systems was particularly challenging because water is channeled through both pipes and open channels. there are points where rain water is ejected from a drainage pipe and travels through the air onto glass panels so the NInsight team needed to work in
Architectural researchers at NInsight were integral in assessing and improving the drainage system of the Florence Railway Stations roof the buildings most striking and challenging feature.
image courtesy Foster + Partners.
NInsight researchers used ANSYS software to create 3-D CFD simulations for a variety of rainfall conditions in assessing performance of the station roofs unique open- and closed-channel drainage system. This simulation shows two different drainage streams (red and blue) converging in a single channel near the perimeter of the building. As shown here, the original design resulted in a large jet of water being ejected from the channel at this intersection. To avoid this, the perimeter channel was modified to include a secondary drainage system for overflow conditions.
multiple phases to study the effects of both air and water. Fortunately, ANSYS CFD software has the capability to model both typical water flows and the more complex open-channel hydraulics associated with the roofs unique design. the NInsight team conducted a number of complex computational fluid dynamics simulations of key intersections of the roofs open-channel drainage system to assess real-world consequences of overflow rainfall. By using ANSYS software to simulate various characteristics of these intersections, NInsight researchers used various flow models to assess the impact of various degrees of blockage of the lower channel. the simulations demonstrated that there were some intersections where water volume would present a danger to pedestrians walking below so the roofs drainage system was modified to address this issue. By adding a second drainage channel underneath the primary system to handle overflow conditions, rainfall was distributed in a way that would protect pedestrians and minimize erosion issues associated with jets of ejected drainage water hitting lower portions of the glass roof. Slight modifications suggested by NInsight have been incorporated into the final design of the Florence railway Station. Analysis with ANSYS software, which was conducted in 2007, allowed the construction project to proceed as scheduled, without the need to build actual models of the drainage system and conduct physical tests with water. n
www.ansys.com
20
TurBOmAChiNEry
Breath of Fresh Air

Cummins uses high-performance computing to deliver better turbochargers to market faster by simulating the entire stage.
By John Horsley, Engineer, Air Handling, Cummins Turbo Technologies, Huddersfield, U.K.
As a developer of turbochargers, Cummins turbo technologies has wrestled with and now resolved running full-stage simulation to ultimately deliver a better product. the enormous size of turbocharger simulation models, averaging 8 million elements for the compressor stage and 12 million for the turbine stage, previously made it impossible to run full-stage simulations as part of the normal design process. ANSYS computational fluid dynamics (CFD) software plays a critical role in the design process by enabling Cummins engineers to evaluate a wide range of design alternatives to optimize performance prior to building a product or even a prototype. Cummins upgraded to a 160-node high-performance compute cluster to take full advantage of the softwares near-linear parallel processing capabilities. the new high-performance computing (HpC) system has reduced solution times by a factor of 12, making it possible to routinely run full-stage simulations. As a result, Cummins has been able to improve turbocharger performance and reduce time to market. Cummins turbo technologies produces turbochargers for diesel
engines used in automotive, offhighway, power generation and marine applications. Now a division of global power leader Cummins Inc., it was founded as Holset engineering, producing a brand synonymous with turboc h a rg i n g a n d a i r f l o w management for the last five decades. turbochargers from Cummins turbo technologies are found in mediumand heavy-duty diesel engines around the world. In simple terms, a turbocharger comprises a turbine and compressor connected by a common shaft supported on a bearing system. the turbocharger compresses air flowing into the diesel engine. this enables the engine to squeeze more air into a cylinder so that more fuel can be added, which, in turn, increases the power of the engine. In the past, Cummins designed turbochargers by going through multiple build-and-test hardware prototype cycles to verify performance, stress and fatigue life. this process was very time consuming and
A typical turbocharger supplied by Cummins Turbo Technologies
ANSYS HPC technology is enabling Cummins to use larger models with greater geometric details and more-realistic treatment of physical phenomena to generate results in less time.
expensive. When the company began converting to an analysis-led design strategy, Cummins turbo technologies selected ANSYS CFX software as its primary CFD tool because of its industry-leading technology for turbomachinery simulation. ANSYS CFX enables Cummins engineers to evaluate design concepts and variations virtually in a fraction of the time and cost required in the past using physical prototypes. While the data that can be captured from hardware prototypes is limited by sensor capabilities, CFD enables engineers to easily determine what is happening at any point in the computational domain. Simulation results show very good correlation with physical testing in the lab on the gas stand.
www.ansys.com
21
TurBOmAChiNEry
Static pressure rise through the compressor stage as well as the level of mesh and flow detail provided by simulation
one of the challenges in simulating turbochargers is the need to use extremely dense meshes to accurately represent the complex geometry and to capture the full flow features of todays advanced turbochargers. this, in turn, creates extremely large problems that require huge amounts of computing power to solve. Cummins previous generation of HpC included eight nodes of parallel computing capacity. Simulation of a full-stage compressor took between 36 and 48 hours, and a full-stage turbine simulation took between 60 and 72 hours. this was too long to keep pace with the design process, so engineers typically simulated only parts of the compressor or turbine stage, such as the turbine wheel, nozzle or housing. though these simulations were valuable, engineers felt that they could achieve their design objectives considerably faster if they were able to simulate the entire stage to better understand the interactions between the various turbocharger components.
to accomplish this goal, Cummins implemented a 160-node highperformance compute cluster running on the linux operating system. most CFD simulations at the company are run on 32 nodes, enabling five simulations to be run at one time. the new system has reduced simulation time by a factor of 12. A single run for a
full-stage compressor now takes four hours, and a full-stage turbine simulation can be run in five to six hours. the improvement in performance is greater than linear: the new generation of hardware is more efficient, and recent improvements in the CFD softwares parallel processing further reduce time. the ability to run full-stage simulations while keeping pace with the design process provides insights that were not available in the past when simulating individual components. engineers can gain a much better understanding of the interaction between the compressor wheel and the diffuser when simulating the entire system. For example, the compressor stage housing is asymmetrical, which results in a variable pressure distribution around the circumference of the compressor wheel. engineers can now easily visualize how that pressure distribution affects compressor wheel performance. using HpC provides Cummins engineers with enough detail in the simulation to fully understand design
Interaction between volute and compressor wheel highlighting pressure variations around the vaneless diffuser
22
www.ansys.com
TurBOmAChiNEry
tradeoffs while obtaining results in a time frame that allows for maximum impact on the design process. HpC also enables engineers to increase me s h d e nsit y t o help im pro ve accuracy of the simulation results. one area in which density is critical is the clearance path around the circumference of the compressor wheel, where air can leak out through tiny gaps. this area is difficult to design because the manufacturing process can introduce variations, causing the gaps to be bigger or smaller than the nominal design. HpC makes it possible to generate a mesh that is fine enough to accurately determine the impact of small variations in these gaps as well as other critical features. engineers can then evaluate the full range of possible gap sizes to determine the impact of different manufacturing methods and tolerances. engineers at Cummins now use full-stage simulations to design new products, validate changes to existing products, and evaluate new technologies. they also evaluate many
different compressor stage designs to see which one provides the best balance between efficiency, flow range and pressure ratio. Customers of Cummins typically are seeking high efficiency, a wide flow range and a specified pressure ratio. these constraints often conflict with each other: For instance, i n c re a s i n g p re s s u re r a t i o generally reduces efficiency. turbine design also requires dealing with many other tradeoffs. For example, the turbine wheel needs to match the flow range while providing the highest possible efficiency and meeting mechanical design constraints. ANSYS HpC technology is enabling Cummins to use larger models with greater geometric details and more realistic treatment of physical phenomena to generate results in less time. Instead of spending months to build and test a physical prototype or days to simulate a design using their previous-generation HpC solution, Cummins engineers can now
Expansion of gas through a variable geometry turbine stage and level of detail defined in the mesh used for simulation
HPC Technology for Large Problems

ANSYS CFX software uses a single-program multiple-data (SpmD) parallel implementation, running identical versions of the code on multiple processors. the overall run procedure is divided into two steps: a partitioning step, in which the mesh is divided into partitions, and a solution step, with a solver process running for each mesh partition under the control of an overall master process. Special multi-pass partitioning is available when simulating coupled rotating and stationary components in turbomachinery applications to optimize the partitioning algorithm and, thereby, improve parallel performance. In general, the solver maximizes the parallelization of numerically intensive tasks and minimizes the serial tasks handled by the master process. Communication between processes during a parallel run is performed using the latest message-passage libraries, including platform Computing mpI and microsoft mpI. Solution scalability is excellent, showing near-linear speedup in many cases.
simultaneously evaluate up to five full-stage compressor or turbine designs in a few hours. physical testing still plays a vital role in the design process, but the number of prototypes that have to be built and tested has been substantially reduced. through extensive simulation, Cummins engineers are now usually able to get the design right the first time. the result is that Cummins can bring new products to market in less time while substantially reducing engineering expenses. the power of HpC has made it possible for Cummins to envision combining both turbine- and compressor-stage simulation with a full-engine simulation to optimize their designs to an even higher level. n
www.ansys.com
23
CONSTruCTiON
Better Cooling, Hot Savings

Fluid dynamics keeps data center server racks cooler using less electricity.
By Marco Lanfrit, Continental Europe Consulting Manager, ANSYS, Inc.
Heat is the enemy of electronic circuits. So engineers at the ANSYS data center in otterfing, Germany, cranked up the air conditioning full blast when peak temperatures near the servers rose above acceptable limits. Hot spots remained, however. engineers used fluid dynamics to identify airflow problems, and the insight it provided led to a quick and inexpensive way to even out room t e m p e r a t u re s a n d s i g n i f i c a n t l y improve cooling efficiency. Hot spots were concentrated near three of six racks that house data center servers and other hardware, such as backup power supplies and data storage units. peak temperatures averaged 41 C (106 F), far above the 34 C (93 F) temperature needed for optimal operation and dangerously close to 45 C (113 F), at which heatsensitive electronic components start to fail. the three hot racks in question house servers used to run the centers solutions for extremely large problems including highly detailed analyses
that exceed customer computing capacity, complex simulations for engineering applications contracted by the ANSYS engineering team, and internal test cases run by ANSYS for physics-related research. these and other complex simulations use high-performance computing (HpC) in which large problems are divided into smaller segments and then solved in parallel, rather than in conventional serial fashion. parallelization allows the solution of extremely large problems by using multi-core Cpus that each contain more than one microprocessor. At data centers such as the one in otterfing, racks of these HpC multicore Cpus may contain hundreds or thousands of microprocessors, giving them the ability to complete complex simulations in a few hours. A major related issue is dissipating heat generated by so many microprocessors running simultaneously. overhead space limitations ruled out using the standard-practice procedure of applying direct cooling
through perforated panels on a raised floor. In search of an alternate solution, engineers created a simulation model representing the major factors affecting thermal distribution in the data center. the room is relatively small, with 32 square meters of floor space and a height of 2.5 meters. Server racks are approximately 2 meters high, leaving about a halfmeter clearance to the ceiling. there is approximately 2 meters of clearance to the adjacent side wall at either end of the rack row. High-power fans blow out hot air, approximately 26 kW of thermal energy, from the three HpC cluster racks containing approximately 450 microprocessors. Another 9 kW of energy exits the other three racks, which contain several midrange conventional servers, a backup server, data storage units, a file server and an uninterruptible power supply. At the rear of the room (behind the racks, an ideal position for sucking in cool air) are two cooling units mounted horizontally near the ceiling.
Server racks viewed from front of room. The three racks on the left house HPC servers. Ideally, cool air from wall-mounted air conditioners on the rear wall (behind the racks) gets drawn into the racks, and hot air from the servers is expelled out the front. Racks to the right of the HPC servers contain mid-range conventional servers, backup server, data storage units, file server and an uninterruptible power supply.
Once thermal partitions were installed in the data center, the room temperature could be turned up to save electricity and still maintain proper cooling for the equipment.
24
www.ansys.com
CONSTruCTiON
each unit is rated at 12.7 kW cooling with an air-handling capacity of 8,000 m3/hr more than enough to sufficiently handle the 35 kW of thermal energy given off by all equipment in the room. Based on these parameters, fluid dynamics simulation using ANSYS FlueNt software produced heat contours for the walls, indicating a hot spot in the corner where hot air exited the HpC server racks. this hot area continued along the top of the adjacent side wall and was apparent at the air inlet for the HpC server. Cooler temperatures existed along the side wall farthest from the HpC servers. temperature measurements in these areas agreed closely with the simulation prediction. A streamline output showing airflow and related temperatures helped to explain the hot spots and uneven temperature distribution in the room. essentially, hot air from the HpC racks (especially the rack at the end of the row) was blowing toward the front of the room and was sucked over and around the racks directly back into the inlet side of the racks, instead of being drawn into the cooling unit as intended. In addition, hot airflow from the HpC server was diverting airflow from the cooling units, so cool air was only partially reaching the HpC racks, thus further raising the temperature of air being drawn into the HpC server racks by fans in the units. With this insight, engineers reasoned that adding two sections of thermal partition (one extending from the top of the racks to the ceiling and another from the side of the HpC racks to the adjacent wall) would resolve these issues. the team quickly evaluated four other configurations using the same room model. though some configurations were marginally more efficient than the one selected, they were ruled out because of greater complexity. Simulation of the chosen configuration (with the thermal partitions
www.ansys.com
Before
After
Contour plot (top) shows hot spots resulting from hot airflow traveling over and around the rack units directly back into the rack inlet. This hot airflow pushes cooling air partially away from the HPC racks to further increase air temperature drawn into the racks (bottom).
Addition of a simple thermal partition significantly improved temperature uniformity and airflow.
added) showed dramatic improvement in delivering temperature uniformity in the data center room and cooling the racks. According to the analysis, HpC rack peak temperature was lowered from 41 C (106 F) to 34 C (93 F). Additionally, increased cooling efficiency in the range of 10 percent to 20 percent enabled engineers to actually raise average room temperature 8 C (approx 15 F) by adjusting the air conditioning control to a warmer setting.
Hot air from rack fans before renovation Air from cooling units
temperatures predicted by the simulation proved to be correct when measurements were taken following partition construction. this improvement, the result of adding two simple thermal partitions, translates into an average annual cost savings of 10,000 euros an outstanding return on an investment for a project that took about two days of work in engineering time and just a few more days to renovate. n
Front wall row of six racks racks containing HpC servers Wall-mounted cooling units blow cool air toward the back of the racks
thermal partition added
rear entry doors
Original data center room configuration
Constructing thermal partitions (highlighted) at the top and one side of the racks was a quick and inexpensive way to modify airflow for correcting the overheating problem.
25
hEALThCArE
Cut to the Bone

Simulation reduces surgery duration and patient risk in treating infants with skull disorders.
By Marek Gzik, Associate Professor, and Wojciech Wolanski, Department of Applied Mechanics Faculty of Mechanical Engineering, Silesian University of Technology, Gliwice, Poland Dawid Larysz, M.D., Department of Pediatric Neurosurgery, Medical University of Silesia, Katowice, Poland
Craniosynostosis is a pediatric condition in which the sutures between the bone plates of an infants skull close before brain growth is completed. the head is often misshapen, but more important is that the condition might lead to developmental difficulties. Craniosynostosis can be treated via surgery to separate the fused structures. Determining where to cut and how many cuts to make is difficult. researchers at the Silesian university of technology in poland are using ANSYS multiphysics software to reduce the number of cuts required for bone expansion. After 20 successful operations, neurosurgeon Dawid larysz of the medical university of Silesia reports that the approach has helped to improve patient outcomes through reduced surgery duration. A newborns brain grows and develops rapidly, doubling in volume in the first nine months and tripling within three years. the skull must expand rapidly to accommodate this growth. the normal infant skull consists of several bone plates separated by fibrous joints called sutures. these sutures respond to brain growth by stretching and producing new bone, allowing the skull to grow in unison with the underlying brain. In time the sutures close, forming a solid piece of bone.
Frontal bones Metopic suture
A child with craniosynostosis requires frequent medical evaluations to ensure that the skull, facial bones and brain are developing normally. The least invasive type of therapy involves use of a form-fitting helmet or band that fits snugly on the prominent areas of the head but allows the recessed and flattened portions to gradually expand into the open areas of the helmet, molding the head as it grows.
Anterior fontanelle Colonal suture
Parietal bones Sagittal suture
Posterior fontanelle
Occipital bone
Lambdoid suture
Sutures in the infant skull
With craniosynostosis, the sutures close too early, reducing expansion and new bone creation, causing the brain to take the path of least resistance. ultimately, the shape of the brain, skull and face becomes distorted. Severe forms of craniosynostosis may result in developmental delays or mental retardation. the condition affects one in 2,000 live births; it affects males twice as often as females. In severe cases, surgery is the recommended treatment. the surgeon flattens the forehead bone to make it more flexible and then makes radial cuts, called osteotomies, in the bone to weaken it, intentionally allowing it to deform as the brain develops. Ideally, surgical correction should occur between the ages of three months and six months. the challenge for surgeons is to perform the minimum number of osteotomies that will allow the brain to grow to its proper size without deforming the skull or reducing its strength. Infant prepared for corrective surgery once the osteotomies have been completed,
www.ansys.com
26
hEALThCArE
the skull should duplicate as closely as possible the mechanical properties of a normal skull. Neurosurgeons rely on knowledge and experience during pre-operative Three-D model of skull before correction planning. their natural tendency is to err on the side of safety, performing enough osteotomies to ensure there is optimal room for brain growth. However, a larger-than-necessary number of osteotomies lengthens surgery time, increasing risk to the infant. Another potential consequence is a permanently weakened skull. Silesian university of technology researchers have pioneered the use of finite element analysis to determine the minimum number and location of osteotomies that are required to enable brain growth. they use computed tomography (Ct) scans to generate a 3-D model of the infants skull with the use of mimics software. this imageprocessing package, developed by materialise NV, generates 3-D models from stacked medical images. the researchers then export the mimics model to ANSYS multiphysics software for analysis. the team determined the bones material properties by performing physical tests using skull material that had been removed in previous operations on an mtS Insight 10 kN test stand. In the recent case of an infant with trigonocephaly, a special type of craniosynos- Finite element model of forehead bone tosis that involves fusion of the metopic suture in the forehead, researchers created 3-D models of the forehead bone in five variants and exported them to ANSYS multiphysics software. these variants included intact
50 45 40 35 Lead [N] 30 25 20 15 10 5 0 0 0,1 0,2 0,3 Extension [mm] 0,4 0,5 0,6
Structural analysis results show load required to deflect bone and deformation of the forehead bone after deflection. Type of Correction
Intact forehead bone Flattened forehead Five osteotomies Six osteotomies eight osteotomies
Deflection (mm)
20 20 20 20 20
Load (N)
52 37 15 10 4
Table of bone deflection after deformation summarizes the structural results.
bone, flattened forehead, five osteotomies, six osteotomies and eight osteotomies. Doctors had previously determined that the bone should deflect 20 mm to accommodate brain growth. using the simulation software, analysts determined the amount of load that was required to achieve 20 mm deformation with each variant. Analysis of the structural results showed that eight osteotomies provided the best skull correction conditions for this patient. Without simulation, doctors probably would have performed unnecessary additional osteotomies to be sure there was enough room for brain expansion. the geometry of each skull is different, so researchers have analyzed the skulls of 20 infants with craniosynostosis. In each case, they were able to determine the minimum number of osteotomies that would provide the needed room for brain expansion. the simulation results helped the surgeon to prepare better for the operation; the procedure itself was much faster, and it was easier on the infant because of the reduced number of osteotomies. operations were successful on each of these patients, and the children are all doing well. the results unexpectedly revealed the important influence of age on correction results. By the time a child is three months old, the skull has stiffened to the point that a substantially larger number of osteotomies is needed to accommodate brain growth. performing the operation at an early age reduces both the number of osteotomies needed and the invasiveness of the operation. this case exemplifies how biomechanical modeling can be used to support surgical procedures. n the utilization of biomechanical and fluid flow modeling tools for surgical planning is growing in many areas beyond pediatric skull surgery, including orthopedic implant placement, cerebral aneurysm procedures and soft tissue repairs. In each of these situations, simulation results are helping to reduce surgery time and provide optimal treatment.
Material Bone of three-monthold child
Young Modulus [MPa] 380
Poisson Ratio 0,2
Peak Load [N] 45
Material properties determined by physical testing
www.ansys.com
OiL ANd gAS
Separating the Streams

multiphase simulation can improve performance of oil and gas separation equipment.
By David Stanbridge, Managing Director, Swift Technology Group, Norwich, U.K.
Separators are used throughout the oil and gas industry to split production fluids into components of oil, gas and water (as well as contaminants). on an offshore facility, the equipment is found in many parts of the overall process. the initial separator, usually referred to as firststage, separates the initial stream into distinct gas, oil and water streams. these streams are then individually processed. poor separation performance can hinder overall production; in some cases, platforms produce only 50 percent of design capacity due to poor separation. the industry has used computational fluid dynamics (CFD) extensively to troubleshoot separation equipment performance with different methodologies. most common is segregated single-phase simulation, in which gas and liquid phases are analyzed separately. multiphase volume of fluids (VoF) simulations are useful in analyzing liquid sloshing behavior in separators secured to moving platforms. this sloshing analysis is usually carried out in combination with a user-defined function that adjusts gravity and applies three inertial forces: Coriolis, euler and centrifugal. Historically, fluids neither enter nor leave the vessel.
Simulation of a vertical cyclone vessel designed to remove bulk liquid from the feed stream. Pathlines of primary gas phase show where the liquid has a concentration of more than 25 percent.
Benefits of Multiphase Simulation As new separation equipment becomes smaller and flow rates exceed the design capacity of existing equipment, end users are questioning the accuracy of both the segregated single-phase approach and VoF for sloshing. extended use of multiphase
28
simulation is now possible as a result of enhancements to computer power and ANSYS FlueNt capabilities. Software improvements have led to reduced run times; multiphase and turbulence models have a greater ability to handle primary and secondary phases. the multiphase method overcomes the limitations of segregated single-phase and VoF approaches. It also allows for detailed analysis of interphase interactions, providing more realistic results. Swift technology Group has studied two types of separation devices that use the multiphase method. the company is a technology-driven organization that offers complete end-to-end product
development for the aircraft, marine, automotive, oil and gas, and renewable energies industries. Droplet separation is fundamental to good separation. the most common equipment for droplet separation is vertical or horizontal vessels that use gravity as the driving force. morecompact separation equipment often uses cyclones. By spinning the flow, employing a standard tangential inlet, or using more-elaborate swirl elements, cyclones can generate accelerations many times that of gravity to potentially provide more efficient separation in a smaller amount of space. However, many other considerations must be investigated. traditionally, cyclonic equipment required exhaustive prototyping and testing to ensure that the many negative consequences were designed out of the final product a lengthy and costly exercise. In a recent r&D program for cyclone development, Swift researchers found that the time for each design change cycle was approximately eight weeks at a cost of around 45,000 (approximately $73,000 u.S.) per cycle, with
Swirl element fitted in the cyclone
www.ansys.com
OiL ANd gAS
seven changes required. By using CFD, each change can be modeled in two weeks, requiring only one actual test saving a total of more than 300,000 (approximately $485,000 u.S.). Note, however, that it is difficult to quantify the exact benefits of simulation in every case.
Simulation of Separation Equipment there are many examples in which the mixture multiphase model has been used to analyze separation within cyclonic equipment. the model is applicable for dilute-to-moderately dense volume loading, for low-tomoderate particulate loading, and for cases in which the Stokes number is less than 1. the simplified model can be used for hydrocyclones equipment whose main function is to separate final oil droplets from water prior to disposal at sea. the comprehensive eulerian multiphase model is applicable to the complex flows that are found in the most common types of separation equipment designed to remove bulk phases as well as re-entrained droplets. users can enhance their analysis of cyclonic flows by applying the reynolds stress turbulence model without limitation for all primary and secondary phases. one important part of separator analysis is commonly overlooked: the impact of upstream piping. this system has a large effect on the distribution of fluids within the vessel. the simulation examples provided horizontal and vertical gravity-driven
separators as well as cyclone-based separators inco rpo rate th e i m pact of u pstream piping. It is difficult to accurately validate the simulation results of the installed vertical cyclone and separator. Simulation has been shown to accurately capture both flow field and separation performance in lab and pilot test rigs. [See references.] using these modeling strategies as well as exhaustive testing performed over many years, all the critical aspects of the flow are correctly resolved and indicate the key performance characteristics. As a result, Swift is changing out the internal components of many vessels based upon simulation results. the main function of a horizontal three-phase separator is to split a feed stream into discrete gas, oil and
Multiphase simulation within a horizontal threephase separator with inlet piping, a vane-type inlet device and full-diameter perforated baffles. The lower layer of fluid is water; above that is the oil phase with the inlet device in the gas phase of the vessel. The pink area at the bottom of the vessel shows where sand entrained in the water phase will initially settle.
The complete length of a typical horizontal separator: The blue layer represents the interface between gas and oil phases, and the green layer represents the interface between oil and water. The vertical blue areas represent part diameter perforated baffles. Along the length of the vessel, four contours show velocity distribution in both oil and water phases.
water streams. Normally, gas is the primary phase, and the two liquid p h a s e s a re s e c o n d a r y. t h e s e liquid phases form droplets that are entrained in the gas phase, and they produce a film on the pipe walls leading to the separator. the first component in the separator is the inlet device, whose primary function is to provide a coarse separation of gas and liquid phases. the gas phase continues along the top of the vessel, while the liquids drop to the bottom of the separator. At the bottom of the vessel, the two liquid phases separate, with the water at the bottom and the oil forming a layer between the water and gas phases. In most cases, perforated baffles are used along the length of the horizontal vessel to control liquid phase flows and to distribute them evenly across the available cross-sectional area of the vessel, minimizing axial velocity and maximizing separation. the eulerian model is required in this
www.ansys.com
29
OiL ANd gAS
Analysis of a vertical production separator with a vane-type inlet device shows that the inlet pipe keeps much of the liquid on one side of the vessel leading to nonoptimal separation.
type of simulation because of the number of fluid regime changes. In a vertical production separator with a vane-type inlet device example, gas and liquid are introduced at the start of the pipe run to the separator vessel. the pipe routing causes the liquid to be biased to one side of the vessel which does not produce optimal separation and, in some cases, can lead to the gross carryover of liquid though the vessels gas outlet. In conclusion, Swift researchers have found that ANSYS FlueNt software can model to a high degree of accuracy many combinations and permutations of separators available within the industry. n
References
Slack, M.D.; Prasad, R.O.; Bakker, A.; Boysan, F. Advances in Cyclone Modelling Using Unstructured Grids. Trans. I. Chem. E., 2000, 78(A), pp. 10981104. Cokljat, D.; Slack, M.; Vasquez, S. ReynoldsStress Model for Eulerian Multiphase. 4th Symposium on Turbulence, Heat and Mass Transfer, 2003, Antalya, Turkey, pp. 8. Slack, M.D. Best Practice Advice for AC3-03 Cyclone Separator, Thematic Network for Quality and Trust in the Industrial Application of CFD (qnet-cfd). Thematic area 3, 2003, Barcelona, pp. 5. http//www.qnet-cfd.net
Helping CAE clients gain insight
IBM High Performance Computing For a Smarter Planet

ibm.com/deepcomputing
30
www.ansys.com
TurBOmAChiNEry
Predicting Turbomachinery Erosion Rates

erosion rate predictions help designers to improve longevity of rotating equipment that must operate in harsh environments.
By Edward Bennett, Director of Fluids Engineering, and Artem Ivashchenko, Project Engineer, Mechanical Solutions, Inc., Whippany, U.S.A.
Solid model of fluid domains of FCC hot-gas expander, illustrating blade characteristics of the machine along with approximate size
turbomachinery flows that contain solid particles represent an undesirable yet often inescapable condition because of practical operating considerations. Dust, sand, fly ash, iron oxide, process-originated materials, and the debris from abradable seals or blade rubbing are examples of the varied composition of these solid particles. these particles can cause erosion, deposition and/or corrosion and are, therefore, a common source of performance degradation and component damage. the consequences are costly, including the obvious expense of turbomachinery components that must be repaired or replaced. Additional costs include penalties from reduced machine-operating efficiency and lost productivity, since the turbomachinery down for maintenance drives the plant equipment. recently, computational methods for erosion simulation have become available, enabling designers to develop moredurable turbomachinery.
mechanical Solutions, Inc. (mSI) is a consulting and r&D company that provides design, analysis and testing services for organizations around the world. mSI, which has proven expertise in design and analysis of all kinds of rotating machinery, has compared erosion rates in two hotgas expander designs.
FCC Expanders Troubled by Erosion Fluid catalytic cracking (FCC) hot-gas expanders form a subset of industrial turbomachinery that consistently sustains significant erosion damage [1]. through the application of powdered catalysts in conjunction with elevated temperature, the FCC process converts highmolecular-weight petroleum hydrocarbons into more valuable petroleum products, including gasoline. the FCC process is often operated continuously in petroleum refineries for periods of up to several months.
CFD models of redesigned FCC hot-gas expander flowpaths for instances of 40 percent reaction (left) and 68 percent reaction (right)
www.ansys.com
31
TurBOmAChiNEry
application of theoretical models developed by Widen tabakoff and team at the university of Cincinnati in the united States [2] [3]. machinery flowpath designers can take advantage of these models to quantify potential damage to passage walls and blades. Furthermore, users can judge the efficacy of proposed design alterations with respect to erosion.
Color contour plots of predicted rotor erosion rate densities for 40 percent reaction geometry (left) and 68 percent reaction geometry (right). Red regions indicate locations of highest erosion rates.
Flue gas, a byproduct of the FCC process, traverses a separator, which removes up to 90 percent of the catalyst particles, then passes through an FCC hot-gas expander. A specialized type of turbomachine, an FCC expander recovers a sizeable portion of the pressure and thermal energy remaining within the flue gas and uses it to drive process equipment or to generate electricity. Beyond having elevated inlet temperatures, singlestage FCC expanders are characterized by a large pressure ratio, typically three to one. the turbine blades of the FCC expander are simultaneously subjected to substantial aerodynamic and thermal stresses that intensify the erosion damage caused by residual solid catalyst particles.
Evaluations of FCC Expander Flowpaths to Reduce Erosion Based on time history data, a particular model of FCC hot-gas expander was found to sustain excessive solid particle erosion damage. While conducting a planned upgrade to a line of expanders, mSI performed a computational erosion analysis of this machines flowpath using ANSYS CFX. the analysis quantified the difference in erosion rates achieved by increasing the design reaction of the expander stage. engineers at mSI designed a new 40 percent nominal reaction flowpath and compared it with another similar also new 68 percent reaction design, for a design pressure ratio of 3.5 and rotor speed of 5,070 rpm. each design consisted of three adjacent fluid domains that represented the stator, rotor and diffuser. Since each domain is cyclically symmetric, only one blade row or sector from each domain needed to be modeled, thus reducing solution time. the models were meshed with ANSYS turboGrid software, producing a hexahedral mesh of approximately 637,000 nodes. the k- turbulence model accounted for turbulence effects, and the stage frame change model was applied to the interfaces between stator, rotor and diffuser domains. the erosion model considers the solid particles size and relative velocity, impingement angle of the particle strike, and materials of both the particle and the eroding surface. the coefficients of quartz steel were used to represent solid particles and eroding surfaces, and the default rebound characteristics of the erosion model were used.
Powerful Analytical Techniques Offer Solutions Computational fluid dynamics (CFD) has been used to optimize the design of turbomachinery flowpaths for many years. ANSYS CFX software, for example, includes the capability to track solid particles in the fluid domain as well as to predict solid particle erosion through
32 ANSYS Advantage 2011 ansys, Inc.
Enlarged color contour plots of the rotor erosion rate densities on blade leading edges of the FCC gas expander are shown for 40 percent reaction geometry (left), and 68 percent reaction geometry (right).
www.ansys.com
TurBOmAChiNEry
the primary erosion damage predicted for the 40 percent reaction flowpath design closely resembled the damage that was sustained by an FCC gas expander of this type operating in a petroleum refinery. this indicated that a good correlation existed between the CFD models and the actual FCC expander. the 68 percent reaction design predicted notably improved erosion performance, especially along the leading edges of the rotor blades. undoubtedly this is due to diminished flue gas velocity on the leading edges of the rotating blades, a characteristic of this design. While both designs feature transonic flow and high aerodynamic loading, mach number plots illustrate considerably reduced stator loading for the 68 percent reaction design. Streamline plots indicate similar flow patterns on the pressure side of each rotor (upper plots). However, simulation predicts a reduction in the hub secondary flow on the suction surface, a performance improvement that resulted from the 68 percent reaction design.
Half-span Mach number color contour plots are shown for 40 percent reaction geometry (top), and for 68 percent reaction geometry (bottom).
Analytical Results Foretold Improvements the predicted aerodynamic performance of the two designs was quite similar, with efficiencies of 87.2 percent and 87.9 percent for the 40 percent and 68 percent reaction designs, respectively. However, differences in the erosion rates were predicted. In switching from the 40 percent reaction to the 68 percent reaction flowpath design, the CFD analysis predicted a 22 percent reduction of average rotor blade erosion rate density along with a 5 percent reduction of average stator blade erosion rate density. For the 40 percent reaction design, the rotor blades had 65 percent higher erosion rate density than the stator blades, but the 68 percent reaction design had only a 35 percent higher rate. this suggested that the rotor and stator blade wear would occur at more consistent rates in the case of the 68 percent reaction design.
Conclusions Fluid flow analysis of two FCC gas expander designs demonstrated the ability to redesign turbomachinery flow paths and to tailor erosion characteristics for improved blade longevity. When compared to the cost of recurring blade replacements, reduced machine efficiency and lost production due to equipment downtime, the cost of addressing erosion with CFD analysis is minimal. Since the extent of the erosion damage caused by solid particles entrained in the primary flowpath can be mitigated, the useful operating life of turbomachines can be extended significantly. n References
[1] Carbonetto, B.; Hoch, G. Advances in Erosion Prediction of Axial Expanders. Texas A&M University Turbomachinery Symposium, Houston, Texas, U.S.A., September 1999. [2] Hamed, A.; Tabakoff, W;. Wenglarz, R. Erosion and Deposition in Turbomachinery. Journal of Propulsion & Power, 2006, vol. 22, no. 2, pp. 350360. [3] Hamed, A.; Tabakoff, W.; Rivir, R.B.; Das, K.; Arora, P. Turbine Blade Surface Deterioration by Erosion. Journal of Turbomachinery, 2005, vol. 127, pp. 445452.
Rotor blade surface streamlines, colored by velocity, near rotor blade surfaces for 40 percent reaction geometry (left pair) and 68 percent reaction geometry (right pair) of the FCC hot-gas expander. Left of each pair represents pressure side and right, suction side.
www.ansys.com
33
ThOughT LEAdEr
Green Design
With simulation and analysis, going green offers many benefits.
By Peter A. Bilello, President, CIMdata Inc., Ann Arbor, U.S.A.
evidence is mounting that the impetus for going green and developing environmentally sustainable products is moving away from mere regulatory compliance to the realization that a significant new business opportunity is at hand. the realization marks a dramatic opinion shift. many u.S. manufacturers greeted sustainability and greenness as a regulatory burden foisted on them by european bureaucrats. today, many of these same manufacturers see that developing sustainable products offers a sustainable competitive advantage. As a result, manufacturers even those subject to little real oversight until recently are undergoing a fundamental attitude adjustment. Corporations realize the green opportunity is much bigger than the emerging technologies wind-powered turbines, solar collectors, biofuels, clean coal, etc. that dominate media coverage of business.
A survey by the massachusetts Institute of technologys Sloan Management Review (winter 2011) and Boston Consulting Group shows that sustainability pressures are changing top managements outlook. Businesses are increasing investment in, and paying more attention to, sustainability; a performance gap is emerging between companies that embrace sustainability and those that do not; and top management is on board more for bottom-line impact than for environmental reasons. essentially, the study looks at reactions to a series of stringent regulations aimed at enforcing sustainability. enacted in the european union (eu), these include the restriction of Hazardous Substances (roHS) and closely related Waste electrical and electronic equipment (Weee) directives. more detailed is the registration, evaluation, Authorization and restriction of Chemicals (reACH) restriction. end of life Vehicle (elV) attempts to reduce the amount of waste when vehicles are finally scrapped. China, Japan, South korea, several other countries, and a number of u.S. states are adopting these regulations or authoring similar ones. the business implications have not been missed in Germany and several other export-oriented economies. recognizing that green confers sustainable competitive advantage and boosts employment, legislators appropriate hundreds of millions of euros annually in loans, grants and subsidies for major industries. Solar heating panels are one example. perhaps the biggest case in point is the automotive industry. manufacturers of cars, trucks and buses are creating radically new powertrains engines, transmissions, fuel cells and batteries plus power electronics for coming generations of electric and hybrid vehicles. Automotive companies worldwide are investing tens of billions of dollars/euros annually.
34
www.ansys.com
ThOughT LEAdEr
Simulation and Analysis Meets PLM Whether in automotive or other large-scale manufacturing, going green brings together two complementary solution sets that, when paired, can significantly enhance an organizations ability to design and deliver green products. Simulation and analysis (S&A) is often called computer-aided engineering (CAe) or computational modeling. Finite element analysis (FeA) and computational fluid dynamics (CFD) are its two most widely used forms. Given the increasing scope and complexity of green requirements, too much is at stake to rely on analyses that are performed late in development. For large, complex green systems, S&A is intended to work within the context of an enterprises overall product lifecycle management (PLM) environment. plm is the enterprise-level solution for managing newproduct development information and all corresponding intellectual property (Ip). plm is a centralized conduit for unified data exchange with product data management and efficient workflow. this approach is as comprehensive as lifecycle sustainability. Success with these powerful solutions and tools means developing a holistic design approach to building green compliance into a sustainable competitive advantage. Significant improvements of product performance and reliability can be achieved with CAe techniques such as parametric CAD modeling, FeA, CFD, electromagnetic modeling and multiphysics computer simulations. But for maximum benefit, they must be used within plm environments applied in a broad sense. this means utilizing a consistent set of business processes and focusing data-creation and data-management tools on the products greenness. S&A, Green Design and Manufacturing From experts and analysts, here are some specifics about using S&A solutions in an overall plm environment to find green opportunities. many of these are not new; some are basic, common business sense. Minimize the use of energy-intensive raw materials by trimming excess weight/mass. use engineering simulation to rethink rules of thumb and old margins for error. materials costs could drop by several percentage points. examples abound: one from outside heavy manufacturing is the plastic bottle. Worldwide, 200 million of them are emptied and discarded every day. each one weighs an ounce or more (roughly 50 grams). reducing their weight by 10 percent could save several thousand tons of plastic per day, plus the energy used to recycle them.
Success with these powerful solutions and tools means developing a holistic design approach to building green compliance into a sustainable competitive advantage.
Maximize the use of eco-friendly materials. Design in recycled materials whenever possible in place of energy-intensive raw materials. Identify and optimize promising new design alternatives with S&A and plm to avoid physically building and testing multiple prototypes. eliminating a single prototype can cut weeks from a product launch schedule. Reduce manufacturing learning curves faster with numerical analysis for likely causes of late-stage engineering changes, errors and rework. profits can be generated sooner. Resolve bottlenecks as they appear in production with S&A and the information in plm repositories. As work-arounds eliminate potential showstoppers, output per unit of energy goes up again, and production costs fall. Use electronic distribution for all factory floor documents work/assembly instructions, for example. the amount of paper (and factory floor trash) can be slashed by thousands of pages a week in a medium-size plant, and many times that in a large one. Focus S&A on shipping and packaging products and supplier materials. Consider moving final assembly, run-off tests and painting close to the customer. For most companies, shipping still offers many ways to reduce fuel costs and get greener. most packaging can become more eco-friendly.
www.ansys.com
35
ThOughT LEAdEr
Address service and support early in the design phase. Simulation is a proven tool for simplifying field-service access to components based on needs for service. Quicker, easier maintenance reduces the customers lifecycle costs and is easily demonstrated with S&A. The end of the product lifecycle brings engineering modeling back to roHS, Weee, reACH, elV, etc. obsolescence and foreseeable recycling difficulties are best dealt with at conception by analyzing what can happen in the scrap yard. most of these green practices have much in common with already-established manufacturing initiatives. these include lean (minimizing inventories), re-engineering (optimizing factory work flows) and sensible risk management. Going green adds to incentives and payoffs; it can leave less astute competitors floundering.
The Green Consumer and Noncompliance Simulation and analysis within an overall plm context offers significant opportunities to win new customers. Consumers have already shown they are willing to pay a premium for well-engineered green products, beginning with hybrid-powered and electric cars. As industry shifts to sustainable goods, todays price differences can be expected to reverse. Failing to comply with sustainability-related regulations can be extremely costly in fines, penalties and recalls. regulators set very low thresholds for certain metals and chemicals; they have low tolerance for even small violations. In some countries, noncompliant products may be banned from sale if a hazardous material threshold is exceeded on a single part. Delays in demonstrating compliance can slow or halt a product launch, potentially leaving products stacked in warehouses while forms are retrieved, material levels are verified, and approvals are sought. S&A in an end-to-end plm environment allows businesses to compile, correlate, analyze, support and report against todays and the futures even-tougher green regulations and requirements.
Green is Here to Stay As a critical part of an overall product lifecycle management environment, engineering simulation provides the data creation and optimization capabilities needed to truly optimize greenness of a new product in all aspects. Industrial companies, regardless of size or line of business, can and will have to utilize plm to integrate green concepts into every phase of product development. manufacturers can speed up and smooth out new design processes and production launches. Alternatives to improve green designs can be quickly and thoroughly evaluated with S&A. engineers can comply with moredemanding governmental requirements, meet changing market needs, boost the value of green products to end users and even optimize conflicting consumer demands. the value of these solutions is that through a simulation-driven product development process, companies can take a proactive approach to gain significant competitive advantage for years to come. they will be able to get green products to market quickly and efficiently. Disruptive last-minute changes can be avoided along with the oppressive costs of noncompliance. New leaders in concerns for environmental issues are already emerging. An overall green approach to product lifecycle management not only enables a company to design, produce and deliver greener products to the market, it also can help the company increase a products perceived value to the consumer decreasing lifecycle costs for consumer and company. plm is truly a green enabler, but a companys plm strategy needs the appropriate S&A tools to turn green and sustainability into sustainable competitive advantage. n
CIMdata is a non-biased, independent, global management consulting firm that has established itself as a world-leading source of information and guidance to both industrial organizations and suppliers of PLM technologies and services.
Simulation images courtesy Cummins, inc. and dar El-handasah. graphics iStockphoto.com/skegbydave, iStockphoto.com/henrik5000, iStockphoto.com/4x6, iStockphoto.com/Falconiaz
36
www.ansys.com
BEST PrACTiCE
Gas turbine supplier leverages ANSYS Workbench to speed complex turbine blade analysis.
By Page Strohl, Senior Structures Engineer, Chromalloy Gas Turbine LLC, Orangeburg, U.S.A.
the extreme geometric complexity of turbine blades and the need to capture this complexity to generate accurate simulation results make turbine blade analysis challenging. the ANSYS parametric Design language (ApDl) for ANSYS mechanical software has been the default e n v i ro n m e n t f o r t u r b i n e b l a d e analysis for many in this industry because of its wide feature set, maturity, and access to powerful scripting tools. But the ANSYS Workbench environment has matured rapidly over the past several releases, and today the software offers major advantages in turbine blade modeling. these advantages include an improved ability to transfer data between analysis applications, new simulationspecific modeling capabilities and meshing advancements.
Gas turbine supplier Chromalloy is gaining significant benefits by moving to the ANSYS Workbench environment. the company is the worlds largest independent supplier o f re p a i r s , c o a t i n g s a n d FA A approved replacement parts for turbine airfoils and other critical engine components for commercial airlines, the military and industrial turbine engine applications. A recent turbine blade analysis in ANSYS Workbench was completed in half the time that the traditional interface would have required. In addition, the ANSYS Workbench model required a much reduced solution time. one key advantage of ANSYS Workbench for Chromalloy is that it enables data sharing among the wide range of analysis tools in the ANSYS portfolio. typically, engineers use
steady-state thermal analysis to determine the thermal state of the blade. the team then brings those results into a static structural analysis and uses the stress-stiffened state for modal analysis. the model created in ANSYS Workbench can also be used for computational fluid dynamics (CFD) simulation with ANSYS FlueNt software. Currently, Chromalloy engineers use in-house tools to generate thermal boundary conditions, but, in the future, they plan to switch to CFD for improved accuracy. the ANSYS Workbench project schematic view saves time by providing an overall view of the entire simulation project. the user simply drags the desired analysis systems from the toolbox and drops them into the project schematic. the analysis systems contain all necessary
ANSYS Workbench schematic of a turbine blade analysis
www.ansys.com
37
image iStockphoto.com/michael1959
Fast Study
BEST PrACTiCE
The face delete function removes unnecessary surfaces in a turbine blade model.
components and guide the user through the system from top to bottom. required data transfer connections are formed automatically. Forming a link using drag and drop in the project schematic achieves data transfer between different physics and creates imported loads in downstream simulation. Simulation often demands modeling capabilities that either are lacking in CAD systems or implemented in a fashion that is not well suited for performing simulation. the ANSYS Workbench environment includes a feature-based solid modeler called ANSYS Designmodeler, which can be used to create parametric geometry from scratch or to prepare existing computer-aided design geometry for analysis. ANSYS Designmodeler software provides a wide range of modeling functions designed specifically for simulation, including parametric geometry creation, CAD geometry modification, automated cleanup and repair, and custom tools designed for fluid flow, structural and other types of analysis. CAD models are usually intended to accurately represent the design intent and often lack additional features needed for simulation. ANSYS Designmodeler makes it easy to define these features. For example, the projection feature projects a line onto a surface to define a local area
38
that can be used to apply a load or add a fine mesh control. CAD models often contain much more detail than required for simulation, and this detail can substantially increase solution times. ANSYS Designmodeler provides a wide range of features for simplifying geometry, such as deleting CAD features, merging multiple bodies into one, and grouping bodies to form a part. Face and edge merge operations easily simplify models by eliminating unnecessary features and boundaries, resulting in faster solutions. Sew and connect operations ensure proper connectivity in models with gaps and overlaps. For example, the face delete function improves meshing efficiency by deleting unneeded surfaces. the merge surface function is used to
merge an island of a surface with a surrounding surface. Boolean operations merge multiple surfaces into one to eliminate unnecessary features and further improve simulation efficiency. Chromalloy engineers use the body operation function in the ANSYS Designmodeler tool to cut out a piece of the model and move it to another area. the hide faces function makes it possible to remove external faces from view, revealing internal surfaces s o th at w or k on th em can be performed easily. the ANSYS Workbench environment offers several major advances in meshing functionality that reduce the time required to mesh complex geometry. meshing technologies within ANSYS Workbench provide
The merge function simplifies a turbine blade model.
www.ansys.com
BEST PrACTiCE
robust, well-shaped, quadratic, tetrahedral meshing on even the most-complicated geometries. users can generate pure hex meshes using several different mesh methods, depending on the type of model and whether the user wants a pure-hex mesh or hex-dominant mesh. Beyond tet and hex meshing, there are extensive tools for hybrid meshing with hexahedral and tetrahedral regions and bodies. these models can be meshed with conforming or nonconforming meshes at common interfaces. the face/edge sizing tool makes it easy to change the size of the mesh. the match control tool applies a mesh on a surface that matches the mesh on another similar surface. the mapped mesh control applies a structured, curvilinear mesh. there are two basic approaches to p e r f o rming a na lysis in ANSYS Workbench. the first approach couples ANSYS mechanical ApDl and ANSYS Workbench mechanical capabilities to perform thermal, structural and modal simulations. C h ro m a l l o y e n g i n e e r s p e r f o r m thermal analysis in ANSYS mechanical
ApDl so that thermal information from an in-house code can be included in the analysis. A dummy thermal analysis is then solved in ANSYS Workbench mechanical to import the temperatures from ANSYS mechanical ApDl. the thermal model is solved from the interpolated loading for conduction alone. these results are then used in ANSYS Workbench as input for static and modal analyses. the second approach is Match control tool used to generate mesh. to perform the analysis c o m p l e t e l y i n s i d e t h e A N S Y S directly linked from thermal simulation. mechanical application driven by Contact between the blade and disk command snippets in the thermal for modal analysis is normally simulation. All the analyses described addressed by attaching contact in the first approach occur in command elements to the disk. to obtain snippets, which are macros modified to stresses, its necessary to run a operate within the ANSYS Workbench separate structural solution with environment. the structural and modal frictional contact. A recent example shows what this analyses are performed in the same way as described in the first approach. functionality means to an individual Structural simulation can be performing gas turbine analysis. the performed with a combination of team at Chromalloy recently created simulation boundary conditions and a turbine blade model in the ANSYS command snippet boundary conditions. mechanical ApDl interface, spending the thermal boundary conditions are a week and a half to generate a meshed model with 5 million nodes. With the ANSYS Workbench environment, it took only four days to generate a model of the same blade with the same level of fidelity but with only 1.5 million nodes. Applying boundary conditions for thermal and structural analysis took three hours on the traditional model, but only 30 minutes with the ANSYS Workbench model. the structural run time for a linear solution without contact elements using the traditional model was one and a half days, compared to only two hours for the ANSYS Workbench model. When contact elements were added, the run time for the traditional model increased to four days, while the ANSYS Workbench model increased to only one day. this application summarizes the substantial benefits provided by A N S Y S Wor k ben ch for tu r bi n e Analysis in ANSYS Mechanical APDL and results interpolated onto an ANSYS Workbench blade analysis. n thermal simulation
www.ansys.com
ANALySiS TOOLS
Warming Up to Direct Modeling

Advanced CAe model preparation enhances pellet stove design and optimization.
By Andrea Tezza, Mechanical Engineering, Extraflame Technical Department La Nordica-Extraflame, Vincenza, Italy
Extraflame Rosy pellet stove with nominal thermal power of 2.5 to 5.0 kW, total efficiency exceeding 87 percent and an hourly pellet fuel consumption of 0.6 to 1.2 kg/hr
Customer preference and restrictive international regulations related to energy efficiency have pushed heating appliance manufacturer extraflame to make significant investments in r&D. la Nordica-extraflame is a leader in the field of heating appliances, with 35 years experience in manufacturing pellet-burning stoves, wood stoves, fireplace inserts and boilers. pellet- and woodburning stoves have gained popularity in recent years as fossil fuel prices soared. Both wood and wood-pellet fuels are carbon neutral by definition: Burning wood does not release any more carbon dioxide than the eventual biodegradation of the unburned wood. In fact, 2.3 kilograms of dried wood provides the equivalent calorific power of 1 liter of diesel oil or 1 cubic meter of natural gas even though wood fuel is priced up to 50 percent less than diesel oil or natural gas. Within these appliances, combustion of pellets or wood increases the temperature of the hot exhaust stream within a plate heat exchanger (built into most models). Air or water is, in turn, heated by the heat exchanger and then circulated throughout the space. extraflame started using fluid dynamics software from ANSYS in 2009 to perform studies on stoves that use wood or wood pellets as the primary fuel source. Analysts conduct detailed component-level fluid flow analysis, such as heat exchanger optimization to improve unit efficiency, which reduces fuel consumption while providing consistent heat. However, the company realized that geometry model preparation was a bottleneck to Simulation Driven product Development. even though most extraflame analysts have some familiarity with traditional CAD systems, access to these tools is not always available. the engineering team recognized that defeaturing a dirty geometry model in traditional CAD can be painstakingly laborious. to better equip analysts for this work and to streamline the process, the department considered both the ANSYS Designmodeler tool and ANSYS SpaceClaim Direct modeler, with the latter selected in 2010.
Whether updating an existing model or designing a new unit from scratch, the CAD department at extraflame provides detailed geometry models for analysts use. these geometries are sometimes shared in native CAD format but, more commonly, the transfer occurs using Step. the ability to import a broad range of geometry formats was an inherent requirement for selection of the analysts geometry tool. Additionally, with a direct modeling tool available that is compatible to ANSYS Workbench, extraflame analysts can build geometries from scratch to facilitate concept studies, prior to any detailed design models being created by the CAD department. When the CAD group provides a detailed design model, analysts perform the defeaturing, cleanup and healing operations, and flow volume creation from solid geometry. this geometry may be adjusted and controlled as part of the optimization process. When a geometry is
Hot air outlet Cast-iron heat exchanger
Pellet loading
Combustion air inlet Room air fan
Exhaust outlet
Detailed CAD model for residential 4 kW (thermal) pellet stove
40
www.ansys.com
ANALySiS TOOLS
To perform flow and optimization analyses, each stove (residential or commercial, wood or wood-pellet fuel source) must be decomposed into two flow volumes: one for the fluid circuit, which circulates air or water to occupant spaces, and one for the combustion circuit, which comprises the combustion air intake and outlet. A plate heat exchanger passes thermal energy from the hot exhaust stream to the fluid circuit.
Hybrid meshes applied to decomposed flow volumes using ANSYS Meshing within the ANSYS Workbench environment
ready for analysis, an associative interface passes the geometry to ANSYS Workbench. meshing, solution and post-processing are all performed within the ANSYS Workbench environment. using ANSYS SpaceClaim Direct modeler, creating the internal flow volumes in a pellet stove is straightforward. the analysis team caps pipes using the Fill command, which then can be dragged using the mouse and the Pull command along the axis of the pipe, automatically creating flow volumes. Dragging the face in the opposite direction extends the pipe volume away from the original domain, which is advantageous at inlets to allow the flow field to fully develop prior to reaching the area of interest. For larger, complex models, the Gaps and Missing Faces commands under the repair tab are helpful for healing problematic geometry. Artificial intelligence built into the GuI takes considerable pain out of the process relative to standard CAD.
Hot exhaust gas stream passing through the plate heat exchanger in the stove
Temperature gradient across the surface of the plate heat exchanger
the Split Edges command is important for eliminating coincident, redundant edges, which commonly exist when working with Step geometry. Small holes, such as bolt holes that need to be closed as part of this process, can be automatically detected and closed using the Fill command. the geometry is simplified by eliminating nonessential parts. the objective is to create two independent flow volumes, one for the exhaust gas stream and the other for the heating medium either water or air, depending upon the unit. these two flow volumes intersect at the stoves plate heat exchanger, which facilitates energy exchange from the hot exhaust stream to the heating medium. the team typically can make analysis-ready geometry within a few hours. Because of the dedicated interface to the ANSYS Workbench platform, analysts can move from geometry to analysis and continue to update the geometry and restart the case without tedious re-work. Following optimization, design changes are provided back to the CAD group so production models may be updated. the annotation utilities in the direct modeler can be used to document final dimensions for reporting purposes. Alternatively, the team may write out the final geometry from the ANSYS product using Step (or other neutral formats) so designers can compare the proposed changes with the initial detailed model directly within the CAD tool. the introduction of ANSYS SpaceClaim Direct modeler along with ANSYS CFX fluid dynamics software tools has had a significant impact on product development at extraflame. Design time has been halved on average. moreover, reliance on physical testing has been significantly reduced, with the number of physical prototypes moving from several to one for a typical project. thermofluid simulation at extraflame now drives and connects the CAD design activity to the products final design. n
www.ansys.com
41
ANALySiS TOOLS
Simulation to Reduce Automotive Emissions

parametric design optimization of selective catalytic reduction systems helps to meet challenges in emission reduction.
By Jayesh Mutyal, Technology Specialist, and Padmesh Mandloi, Senior Technology Specialist, ANSYS, Inc.
tighter diesel emissions standards are being put in place by regulators throughout the world. For example, the united States environmental protection Agency (epA) implemented diesel emissions standards for 2010 that force a dramatic reduction in discharges of particulate matter and nitrogen oxide (Nox). Current stateof-the-art in-cylinder solutions have fallen far short of achieving these limits. Selective catalytic reduction (SCr), which uses a catalyst to convert Nox to nitrogen and water, is one exhaust aftertreatment method that is being considered to meet emissions requirements. Several design challenges stand in the way of achieving the full potential of SCr technology. the most important is ensuring that a given SCr system design achieves the required level of No x reduction over the full operating cycle of a specific engine. SCr performance needs to be evaluated over a wide range of conditions, ranging from traveling at 70 mph down an interstate freeway to delivering goods in stopand-go traffic in the city. A common problem that prevents SCr from
Simulation of total deformation in the exhaust system (left), von Mises stresses plotted on the surface of the exhaust manifold (center), and first-frequency modes (right)
U.S. EPA Standards for HD Diesel Engines achieving its potential is 6 the release of unreacted EPA 1994 5 ammonia, which is called ammonia slip. this can EPA 1998 4 occur when excess 3 ammonia is injected or EPA 2004 2 when there is insufficient EPA 2007 catalyst surface area. 1 these challenges are EPA 2010 0 being addressed by 0 0.02 0.04 0.06 0.08 0.1 0.12 simulation tools from Particulate [g/HP-hr] ANSYS that can be used European Emission Standards for HD Engines to model the fluids, 8 thermal, structural and Euro 2 7 chemical behaviors of 6 SCr components and Euro 3 5 systems. engineers can 4 Euro 4 use computational fluid 3 dynamics (CFD) techEuro 5 2 nology to understand the 1 Euro 6 mixing of urea with exhaust 0 gases, its evaporation 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 Particulate [g/Kw-hr] and decomposition, the On-Highway European and U.S. EPA Standards ensuing chemical Source: u.S. EPA and European union emission standards reactions, and the resultant thermal behavior of exhaust gases Catalytic conversion occurs at and mechanical components. the high temperatures typically above team can apply finite element analysis 500 k. these temperatures are (FeA) technology to model the applied to exhaust aftertreatment structural behavior of components components. Changing engine-load under stress and vibrations. conditions lead to fluctuating exhaust gas temperatures that, in tur n, produce high thermal stresses and the potential for thermal fatigue in exhaust components. Fluidstructure interaction (FSI) technology has been used to apply the wall temperatures calculated by the CFD simulation as an input to FeA, which predicts the stress and deformation of the SCr structure. the CFD and FeA solvers are in the same ANSYS Workbench environment under one platform, allowing seamless data exchange. NOx [g/Kw-hr] NOx [g/HP-hr]
42
www.ansys.com
ANALySiS TOOLS
SCR simulation on a parametric mixer model
Ammonia mass fraction at the catalyst inlet, NH3 uniformity and pressure drop as simulated for mixer design variations
Detailed de-No x reactions are modeled to predict Nox conversion ratios. the results of the simulation can be presented as graphical images and animations that enable engineers to gain a much better understanding of the proposed design than could ever be achieved with physical testing, since the breadth and quantity of information is much greater. For example, the fluid dynamics simulation predicts uniformity of flow, ammonia and isocynic acid at the catalyst entry. Simulation generates animations that plot the flow of particles through the SCr system. thus, the engineering team is able to easily spot dead zones and gain an understanding of the geometrical features that cause them. typically, an original SCr model is based on an existing prototype, and results are then correlated to physical testing to check accuracy. once the model has been validated, a user can quickly evaluate the performance of a large number of design variations for optimization purposes. Bidirectional CAD connectivity, automated meshing, a project-level update mechanism, pervasive parameter management and integrated optimization tools deliver great productivity benefits in analyzing multiple design variations. From within the ANSYS Workbench project window, a series of design points can be easily built up in a tabular form and automatically executed to complete multiple what-if studies of alternative design possibilities. For example, a user can study the effect of injection parameters, such as the
start of injection or injector orientation over a range of values. the ANSYS DesignXplorer tool provides the ability to perform design of experiment (Doe) via response surface methods (rSms) to drive the design process. Doe/rSm can be used to develop experiments that examine many effects simultaneously with relatively few simulation runs. the engineer can iterate to a globally optimized design with a far higher level of certainty and in much less time than with the traditional approach.
the ability to accurately model the performance of SCr design concepts without having to build a prototype makes it possible to evaluate many more designs in the same time frame. Simulation provides even more design data than physical testing, so it can result in substantially improved performance. lower cost and shorter lead times of simulation provide faster time to market and reduced development costs. n
Aftertreatment Utility
Aftertreatment custom utility
ANSYS provides a ready-made custom utility that can be used to dramatically reduce the amount of time required to set up the simulation: the user enters all or most of the needed information using a single screen. once the SCr systems geometry is imported into the simulation environment, it takes only five to 10 minutes to provide the information necessary to perform the analysis. Another
advantage of this semi-automatic process is that it provides a consistent process flow that h e l p s t o p re v e n t m i s t a k e s . once these inputs are entered, the software models the flow region starting upstream of the urea doser and progressing to the outlet of the catalyst. multispecies analysis accounts for each constituent of the exhaust stream.
www.ansys.com
43
AcAdemic
Virtual Speech
Researchers use virtual tools to model and simulate the human vocal tract and upper airway.
By Sidney Fels, Professor, Human Communication Technologies Lab, Department of Electrical and Computer Engineering, University of British Columbia, Canada
Computer-generated models consistently help research groups to understand, diagnose and treat disorders of human oral and airway anatomy such as swallowing disorders due to stroke, chewing disorders after cancer surgery, obstructive sleep apnea (OsA), and various speech pathologies. The dynamic models of these anatomical components for biomedical applications are valuable research tools, but they are often difficult to construct. They require unusually high fidelity and stable simulations of complex, hybrid models. The complex physics includes coupled rigid bodies, finite element method (FEM) and fluid simulations, all in a fast, interactive simulation environment. In addition, the models must be made available in modules that allow groups of researchers to collaborate on complex projects. Frances Grenoble Images Parole signal Automatique lab (GIPsA-lab) and Techniques de llngnierie Mdicale et de la Complexit (TIMC-IMAG) laboratory have used Ansys simulation software successfully for almost 10 years to study speech production mechanisms and computer-aided surgery. These labs jointly developed finite element models of the human tongue, jaw and face. Their models are used to study speech gesture controls and the consequences of maxillofacial surgeries on facial mimics and speech production. More recently, the University of British Columbia (UBC) in Canada extended and implemented the models into its own biomechanical simulation tool, Artisynth, which was developed at UBCs Human Communication Technologies (HCT) lab for human anatomy studies. UBC has been using a combination of Ansys Academic Research simulation software and Artisynth to push the limits of knowledge about human oral and airway anatomy dynamics (the oral, pharyngeal and laryngeal [OPAL] complex) in biomedical and speech applications. One project models the OPAL complex for speech and computer-aided surgery and can provide a goldstandard result to compare with Artisynth, the human biomechanics simulation engine; a second simulates airflow through the OPAL cavity specifically, aero-acoustic simulation of burst (pa) and fricative (sh) sounds.
Structural Dynamics To accomplish the objective, the GIPsA and TIMCIMAG labs generated a 3-D biomechanical model of the tongue and oral cavity controlled by a functional model generating muscle forces coupled with an acoustic model. For example, the diagram in Figure 1 shows how the tongue muscles work when uttering French vowels. Researchers modeled the tongue using ANSYS SOLID185 hexahedral elements with the hyperelastic properties of the five-parameter MooneyRivlin model. The muscles, represented by specific subsets of elements, generate forces to deform and move the tongue and mandible.
Tongue
Mandible
Hyoid Bone
Figure 1. Combined functional model and acoustical model control a 3-D biomechanical model of tongue and oral cavity (top); element subsets generate forces to move the tongue and mandible (bottom).
44
ANSYS Advantage 2011 Ansys, Inc.
www.ansys.com
AcAdemic
Figure 2. Manually designed 3-D finite element mesh consists of three layers of full and degenerated ANSYS SOLID185 hexahedral elements (left); round-shaped lips with a protrusion make the sound ou (right).
The GIPsA and TIMC-IMAG labs developed a 3-D biomechanical model of the human face and focused on lip deformations, since this articulator has a major influence on the acoustic signal generated by airflow coming from the lungs. Using computed tomography (CT) data, the team manually constructed a 3-D finite element mesh, consisting of three layers of full and degenerated ANSYS SOLID185 hexahedral elements. A hyperelastic MooneyRivlin constitutive law accounts for the nonlinear behavior of facial tissue, and muscles fibers are represented by piece-wise uniaxial tensile elements. The project simulated a large number of facial movements and facial mimics that develop during speech, such as lips that form a round shape with a protrusion, which is required to make the sound ou (Figure 2). The GIPsA and TIMC-IMAG labs coordinated their efforts with the modeling work performed at UBCs HCT lab. The lab combined the face, jaw, hyoid and tongue into a rigid-body/FEM framework that provided features for these structures. The features included collision detection and handling, a tightly coupled FEM/rigid body solver, various FEM types, muscle models, rich graphical user interfaces and interactive simulation rates for testing what-if scenarios. Team members worked closely with the results from Ansys software and were able to confirm that the interactive simulations using Artisynth are consistent with the gold standard that Ansys simulation software provides. Effectively, stress and strain computed by Artisynth compared favorably with those from the gold standard provided by Ansys. The lab now has a tool that can be trusted for clinical researchers to investigate the dynamics of the OPAL complex including chewing, swallowing and speech production.
speech. speech movements and acoustic signals are influenced by communicative linguistic goals, perceptual constraints and physical properties of the speechproducing apparatus. To learn how these different factors combine and interact with each other, researchers need an efficient approach that generates realistic physical models of the speech-producing system. In preliminary research, UBC used Ansys FLUEnT software to simulate the air flow of two speech utterances. The first utterance was a fricative consonant, a sound that is produced by turbulence typically due to a channeled flow of air striking the teeth (sh in show). A compressible flow was simulated so the acoustic (pressure) waves could be directly measured, and large eddy simulation (LEs) was used to model the turbulence. A buffer zone around the outlet prevented the waves from reflecting back into the domain. Although the spectra calculated from the simulation results did not agree with experiments on a number of spectral features, they are promising because the amplitudes and trends of the spectra do agree. For example, observe the acoustic waves escaping from the airways mid-sagittal section (Figure 4). The most useful measurement was a direct resolution of the acoustic waves by measuring pressure.
Fluid Dynamics UBC scientists also studied the fluid dynamics of the airflow, which is crucial for speech and plays a role in OsA. In studying speech production, the objective is to understand the complex motor actions that produce
www.ansys.com
Figure 3. Models of the face, jaw, hyoid and tongue are combined in anatomical modeling framework.
45
AcAdemic
high-speed video and particle front simulation, and between microphone experiments and pressure front simulation (Figure 7). Ansys technologies have been a valuable tool in UBCs investigations of oral, pharyngeal and laryngeal anatomy. UBC continues to use these tools to better understand the regions of interest and to help create a tailored clinical research tool to let non-experts more easily access modeling results. n
The author acknowledges other researchers who contributed to this article:

Figure 4. Acoustic waves escape from the mid-sagittal section of airway.
8 4 Height (cm) 0 -4 -8 -5 0 5 10 15 20 25 30 35
University of British Columbia, Canada Peter Anderson, HCT-lab, Department of Mechanical Engineering Sheldon Green, Department of Mechanical Engineering John Lloyd, HCT-lab, Department of Mechanical Engineering Ian Stavness, HCT-lab, Department of Electrical and Computer Engineering Centre National de la Recherche Scientifique (CNRS), France Yohan Payan, TIMC-IMAG Mohammad Nazari, GIPSA-lab Pascal Perrier, GIPSA-lab Florian Vogt, GIPSA-lab
For more information
Width (cm)
Figure 5. Simulations show good correlation with microphone experiments and high-speed videos from actual speech.
University of British Columbia: http://www.ece.ubc.ca/~hct/research GISPA-lab: http://www.gipsa-lab.inpg.fr TIMC-IMAG: http://www-timc.imag.fr ArtiSynth: http://www.artisynth.org
Researchers also simulated a bilabial plosive (pa in paper), which develops when the lips rapidly open to release a pressure pulse from the mouth. The ensuing burst of turbulent air is of interest because it produces a negative microphone pop, and it has a positive influence on speech perception. The team modeled the mouth as a wide oval, using LEs to simulate the utterance as a burst of pressure from the mouth. The simulations agreed well with experiments (Figure 5), demonstrating that a fluidstructure interaction simulation involving lip motion would be needed to capture further complexities of the air flow. Evidence of this close correlation is shown (Figure 6) between the
References
Stavness, I.; Lloyd, J.; Payan, Y.; Fels, S. Coupled HardSoft Tissue Simulation with Contact and Constraints Applied to JawTongueHyoid Dynamics. International Journal of Numerical Methods in Biomedical Engineering, 2011, vol. 27, pp. 367390. Derrick, D.; Anderson, P.; Gick, B.; Green, S. Characteristics of Air Puffs Produced in English Pa: Experiments and Simulations. Journal of the Acoustical Society of America, 2009, vol. 125, pp. 22722281. Buchaillard S.; Perrier P.; Payan Y. A Biomechanical Model of Cardinal Vowel Production: Muscle Activations and the Impact of Gravity on Tongue Positioning. Journal of Acoustical Society of America, 2009, vol. 126(4), pp. 20332051. Nazari, M.A.; Perrier, P.; Chabanas, M.; Payan, Y. Simulation of Dynamic Orofacial Movements Using a Constitutive Law Varying with Muscle Activation. Computer Methods in Biomechanics and Biomedical Engineering, 2010, vol.13(4), pp. 469-482.
40
40 mic experiment pressure front
30 Distance (cm) Distance (cm) normal puff normal simulation 0 0 50 100 Time (ms) 150
30
20
20
10
10
0 0 50 Time (ms) 100 150
Figure 6. Comparison between high-speed video and particle front simulation
Figure 7. Microphone experiments correlate well with pressure front simulations.
46
ANSYS Advantage 2011 Ansys, Inc.
www.ansys.com
Tips and Tricks
Understanding Actuators
Multiphysics system simulation provides insight into the operations of actuators.
By Mark Solveson, Application Engineer, ANSYS, Inc.
Frame
Electromagnetic actuators, or solenoids, are devices that operate by producing magnetic fields to move an armature a desired distance at a desired force. They are used in applications such as fuel injectors, power distribution interrupters and breakers, for example and various automotive, hydraulic and industrial applications. A typical electromagnetic actuator consists of a multiturn coil configured around a ferrous pole and a moveable armature. Additional ferrous parts, such as a frame, provide a return path for the magnetic flux. When the actuator is connected to a voltage source, current flows through the coil and creates magnetic flux in the device, and, as a result, magnetic force is produced that moves the armature from open to closed position. Other actuator configurations may include permanent magnets to assist in flux production or to help hold the armature in place while voltage is switched off in the coil. Voice coil actuators use permanent magnets to produce a magnetic flux that impinges the coil current and, thus, produces a Lorentz force on the coil. These devices may be 2-D or 3-D in nature and can include rotational motion or noncylindrical rotation (rocker motion). T h e f e r ro u s m a t e r i a l o f t h e armature, pole and frame is modeled with nonlinear BH curves to capture
Pole
Coil
Armature
Three-dimensional and axis-symmetric view of typical electromagnetic actuator, with frame, pole piece, coil and armature defined
important saturation effects that can limit performance. The size and shape of the armature and mating pole affect the force profile on the armature and/or the closing time. Additionally, the design of the coil determines electrical resistance and strongly affects coil inductance, since the inductance is equal to the square of the number of coil turns multiplied by the total magnetic permeance of nonlinear ferrous objects and air gaps. The ratio of the resistance divided by the inductance (L/R) is the electrical
Current vs. Voltage, Turns, Time
1.00
Current (Winding1) [A]
0.80
0.60
0.40
Curve Info
Magnetic flux density shown during a transient simulation after 0.001 seconds. Figure shows detail of magnetic diffusion due to eddy currents. As time progresses, the fields diffuse through the thickness of the device, the force increases, and the armature closes once the magnetic force overcomes spring and load forces.
0.20
BatteryVoltage='175V' Turns ='3000' BatteryVoltage='175V' Turns ='3300' BatteryVoltage='175V' Turns ='3600' BatteryVoltage='200V' Turns ='3000' BatteryVoltage='200V' Turns ='3300' BatteryVoltage='200V' Turns ='3600'
0.00
20
40
Time [ms]
60
80
100
Parametric sweep of a current transient versus coil turns and voltage are distributed over several CPUs to reduce total simulation time.
www.ansys.com
47
TiPS ANd TriCkS
time constant; it determines how fast current can rise in the coil. Considering this rise time of current due to the electrical time constant, the magnetic diffusion time (how fast the magnetic flux builds up in the device due to eddy currents) can also affect actuator performance. During a fast rise in current, the magnetic flux is crowded near the inner surface of the actuator before diffusing through the device, thus delaying the buildup of force on the armature. Similarly, magnetic effusion occurs when the voltage is switched off and magnetic flux dissipates out of the device, which can delay the opening of the armature. two- and three-dimensional magnetic analyses can be performed using the static or transient solvers in maxwell electromagnetic field simulation software. often, the coil design (shape factor, number of turns and wire size) and geometry optimization can be done in a series of static simulations in which the current and position are varied to produce a family of curves representing armature force versus position and current. Since maxwell uses automatic adaptive meshing for each of these variations, it is very easy to perform parametric sweeps or optimizations on these variables. Beyond the static simulation approach, full electrical and mechanical transients using the maxwell transient solver can be considered to determine how fast the armature reaches the closed position. A robust simulation of this type can be completed using an arbitrarily defined voltage source (or attached circuit using the maxwell circuit editor), nonlinear materials, and mechanical equations of motion (including damping and load forces, which can be functions of position, speed or time), in which eddy currents and magnetic diffusion are considered. parametric sweeps, design of experiments, or optimization of static or transient cases can be automatically distributed on a cluster of computers or Cpus to reduce total solution time.
SourceVoltage
SourceCurrent_Probe
TurnOnSwitch
Zener
VCurrentDiode
VCoilCurrent
Winding CoilResistance
Coil
HystControlSwitch
HystEngageSwitch
CoilVoltage
10.0
1.00
Armature Position, Coil Current, Voltage
400
8.0 0.75 Armature Position [mm] 300
Coil Current [A]
0.50
200
4.0
.025 2.0
Curve Info Coil Voltage Armature Position Coil Current
100
0.0
0.00
25
50
75
100 Time [ms]
125
150
175
200
Created using the circuit editor within Maxwell, a transient simulation with a chopped current controller shows transient waveforms of position, coil current and coil voltage.
If a more detailed electronic circuit is required, then Simplorer may be used to cosimulate with maxwell. Simplorer software is a multi-domain circuit and system simulation tool that models circuits, block diagrams, state machines and VHDl-AmS components. It also dynamically couples to several other simulation tools from ANSYS.
Robust system simulation in Simplorer shows a drive circuit coupled to the Maxwell FEA model through a coupled transient link. Mechanical pins are connected to define mass, forces, springs and limit stops. Plots show coil current and voltage along with armature position and force versus time.
48
www.ansys.com
Coil Voltage[V]
6.0
TiPS ANd TriCkS
Alternatively, Simplorer software can use an equivalent circuit generated from a parametric sweep of position and coil current (in which eddy effects are ignored). Detailed semiconductor models and closed-loop control systems can be used with a detailed 2-D or 3-D actuator model that can be connected to a mechanical/hydraulic load using the available multidomain components in Simplorer. once the time domain coil and core losses are determined in the electromagnetic models, they can be mapped to ANSYS mechanical software or ANSYS CFD (fluid dynamics) products for thermal analysis within the ANSYS Workbench environment. the time-averaged losses are spatially mapped to the thermal models, in which heat transfer coefficients are assigned or solved explicitly using ANSYS CFD. the steady-state temperature of the actuator is evaluated along with transient thermal performance and thermal cycling. ANSYS provides a comprehensive analysis approach for electromagnetic actuators. Whether the application is static, transient, thermal or system analysis, there is an integrated tool set available to cover any level of analysis. n
ANSYS Workbench enables mapping of the losses from a magnetic simulation to a static and/or transient thermal simulation.
Windows. Life without Walls . HP recommends Windows 7 .
Powering
with the HP Z800 Workstation
with Intel Xeon processor
Visit www.hp.com/go/ansys for more information

2010 Hewlett-Packard Development Company, L.P. The Intel Logo, Xeon and Xeon Inside are trademarks of Intel Corporation in the U.S. and/or other countries. Microsoft and Windows are U.S. registered trademarks of Microsoft Corporation.
www.ansys.com
49
ANSYS, Inc. Southpointe 275 technology Drive Canonsburg, pA u.S.A. 15317

Send address corrections to [email protected].

Revista ANSYS

Uploaded by

Copyright:

Available Formats

Revista ANSYS

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Revista ANSYS

Uploaded by

Copyright:

Available Formats

VOLUME V ISSUE 2, 2011

EXC ELLENCE IN EN G I NEE R I NG SI M U L AT IO N

BUILDING ON A GLOBAL REPUTATION

REDUCE DESIGN TIME

Intel Cluster Ready. Intelligent!

Moving Toward Sustainable Wastewater Treatment

Stepping on the Gas

Forecasting Underwater Noise

Simulation soundly predicts hydro-acoustics during offshore pile driving.

Building on a Global Reputation

Breath of Fresh Air

Better Cooling, Hot Savings

Cut to the Bone

ANSYS Advantage Volume V, Issue 2, 2011

28 OiL ANd gAS

Separating the Streams

Predicting Turbomachinery Erosion Rates

With simulation and analysis, going green offers many benefits.

Warming Up to Direct Modeling

Simulation to Reduce Automotive Emissions

Multiphysics system simulation provides insight into the operations of actuators.

About the Cover

ANSYS Advantage 2011 ansys, Inc.

ANSYS, Inc. | Southpointe | 275 technology Drive | Canonsburg, pA 15317 | u.S.A.

ANSYS Advantage Volume V, Issue 2, 2011

a) Milk volume fraction in prototype 3: a) t = 39.3 s, b) t = 51.9 s

Temperature in prototype 3: a) t = 39.3 s, b) t = 51.9 s

Temperature in prototype 4: a) t = 39.3 s, b) t = 51.6 s

ANSYS Advantage 2011 ansys, Inc.

Moving Toward Sustainable Wastewater Treatment

Treated effluent out

ANSYS Advantage Volume V, Issue 2, 2011

0.9 Normalized DO Obtained from CFD

0.500 0.600 0.700 0.800 0.900 Normalized DO Obtained from Measurement

1 0.8 Normalized DO 0.6 0.4 0.2 0

ANSYS Advantage Volume V, Issue 2, 2011

ASIC: Driving the oscillator signal-processing calibration

Coils: Transmitting coil Receiving coil Reference coil

(CR) (RM) (AM)

Induction sensor architecture

ANSYS Advantage 2011 ansys, Inc.

Background image iStockphoto.com/Okea

Stepping on the Gas

Electronic throttle control pedal position sensor

Circuit simulation setup

198.40 198.60 Time [us]

100.00 Time [us]

Output vs. Angle 5 4 3 2 1 0

Circuit simulation output shows the transfer function of a proposed design.

ANSYS Advantage Volume V, Issue 2, 2011

Original coil design (left) and improved design (right)

Adding a hole to the bracket reduced its effect on output voltage.

ANSYS Advantage 2011 ansys, Inc.

Rendering of steerable conductor

ANSYS Advantage Volume V, Issue 2, 2011

ANSYS Advantage 2011 ansys, Inc.

Stresses on radial locking pads that hold HDH in place

HPC Expedites the Design Process

ANSYS Advantage Volume V, Issue 2, 2011