Optimization of Heat Management of Vehicles Using Simulation Tools

Transcription

1 Seoul 2 FISITA World Automotive Congress June 12-15, 2, Seoul, Korea F2H246 Optimization of Heat Management of Vehicles Using Simulation Tools Rudolf Reitbauer, Josef Hager, Roland Marzy STEYR-DAIMLER-PUCH AG, ENGINEERING CENTER STEYR A-4 Steyr, AUSTRIA The target was to develop a coupling between simulation methods that allow the user to cover overall vehicle heat management analysis. An efficient analysis is only possible when interactions between subsystems of a vehicle are considered simultaneously. As simulation software is available for many subsystems, it is useful to combine them in a way that only less manual iteration steps have to be performed. The subdivision into 3 subsystems, the engine cooling circuit, the cooling air flow and the air conditioning circuit can be covered by efficient and accurate one-dimensional network simulation software tools. For detailed flow information 3-dimensional CFD tools can be brought in. To hand over simulation results from one software package to get boundary conditions for the other package is the first step. To exchange this information in both directions and to consider interdependencies between the subsystems is the second step. In the following typically applications of such software integration methods are shown. Keywords: heat management, simulation, software interfaces INTRODUCTION Considering economical and environmental aspects and rising demands for performance and comfort placed on modern day vehicles are leading to more and more complex cooling systems. The number of possible variants of cooling systems is growing progressively with the increasing number of cooling system s. As an optimization based on tests will consume too much time and money, it is necessary to use simulation tools. When creating a comprehensive simulation model all s relevant for heat management have to be considered. To meet future requirements for covering the overall thermal heat management new simulation models and interface techniques were developed. A very efficient heat balance simulation tool (KULI) based on network systems runable on a standard PC will be presented. Using one dimensional methods and mesh flow theory for modeling is a very efficient method to analyze the cooling air flow. For simulating the system of the air conditioning circuit a module for the above software was developed (KULI AC) It allows to combine the air conditioning system with the engine cooling system in a simultaneous way. The impact of refrigerant phase changes on the heat exchange within the s can be considered by using an advanced 2- phase modeling technique. Another example of such a simultaneous software integration is the interface between a software for the simulation of fluid circuits (FLOWMASTER) and the above mentioned heat transfer software. It allows to simulate engine cooling systems under both points of view, the air side and the fluid side. developed to include data of CFD analysis (STAR CD) into the one dimensional air flow simulation model. Other software as i.e. engine simulation and driving simulation will be integrated in the near future to serve the demands of complete software clusters. Especially when transient behavior must be analyzed integration of these additional subsystems is important. ENGINE COOLING AND TRANSIENT BEHAVIOR The first subsystem, the standard engine cooling system is an object of heat management simulation since many years. The network based air-flow and heat exchange simulation tool KULI was developed further to hit some additional demands. To cover also the fluid flow the software was connected to the software FLOWMASTER. This linkage offers an efficient software duet for simulating the overall engine cooling system within one analysis run. One advantage of this connection is, that transient processes can also be determined. KULI For complex flow patterns however it is desirable to use CFD analysis. To serve this demand a procedure was 1

2 The transient behavior should not only contain the time dependent changing of boundary conditions as vehicle speed and engine operating points, than also the heating up of s and fluids. The following application should demonstrate the possibilities of such investigations. APPLICATION 1 For performing transient calculations of the engine cooling system the user can define time dependent boundary conditions. He has to define the start- and endtime of the period he wants to simulate and the timestep increment for the points where he wants to have analysis results. Basic boundary conditions for this time-window are: engine speed, the according water pump speed, oil pump speed and i.e. thermostatic opening angle. These information can be defined within the FLOWMASTER software. From the entered values together with pump curves the flow rates can be gained as shown in the diagram below Coolant Flow Rates Timet [s] Coolant Flow Radiator Coolant Flow Bypass Oil Flow These flow rates together with the heat rejection of the engine (from a engine map), driving conditions of the vehicle and the heat capacities of the materials (masses) prescribe the simulation problem. Now when running the two software tools KULI and FLOWMASTER simultaneously the one calculates the fluid side and the other the air side and the heat exchange. They are transmitting data during the analysis until both sides have balanced their energy equations. As result the user gets time dependent temperature curves of the various coolants. When looking to the heat rejection diagram it can be seen, that after some time (i.e. 15 [s]) the engine heat is going to the heat exchangers, before this time most of the heat is used to heat up the heat capacities (masses) Heat Performance Time [s] This simple example should show the principle possibilities for simulating transient problems. There is no limitation about complexity of systems. For instance the engine could be modeled up to very sophisticated models, the cooling system himself could be very complex and also the driving conditions can be quite tricky in the points of interest. AIR CONDITIONING IN COMBINATION WITH THE ENGINE COOLING The KULI base module for engine cooling simulation was enlarged with a module KULI AC that enables the user to analyze the complete air conditioning system. The interaction between engine compartment air flows, engine cooling system and climatization of the vehicles interior can be simulated in one loop. KULI AC Radiator Oilcooler Heat Source Temperatures Time [s] Radiator Entry Radiator Exit Oilcooler Entry Oilcooler Exit All present-day used AC s as evaporators, condensers, pipes, knees, expansion valves, compressors and storage tanks are provided by simulation models. It is possible to analyze closed AC-circuits also then single separated s. Boundary conditions can be chosen from the users available data and demands. To consider the phase changes of the refrigerant, an enhanced twophase model with the according heat exchange influences is prepared for getting accurate results. 2

3 Heat Exchangers The measured data is stored in the form of a notebook which contains different pages with tables and buttons. The measured values are transformed into dimensionless numbers for example Reynolds number and drag coefficients. For heat transfer calculation (and also for the calculation of the separated condensate mass flow) the heat exchanger is divided into discrete area elements with their corresponding air mass flow. The change of the air properties and refrigerant conditions in a single element is described by a differential equation. Compressor The compressor model is based on measured data from the volumetric efficiency and the isentropic efficiency. The volumetric and isentropic (adiabatic) efficiency are functions which depend on speed and pressure ratio of the compressor. It is possible to simulate internal and external controlled compressors whether with fixed or variable compressor capacity. Simulation Results At the end of the simulation the results are printed to an ASCII file. The results include the following data of the air conditioning system: Pressures (entry, exit, air, refrigerant) for each Temperatures (entry, exit, air, refrigerant) for each Refrigerant Enthalpy (entry, exit) for each Mass flows (air, refrigerant) for each Refrigerant vapor quantity (entry, exit) for each Heat flows for Evaporator and Condenser Condensate separation at the evaporator surface Surface temperature of the evaporator so that risk of icing up can be estimated Required power, pressure ratio, adiabatic and volumetric efficiency of the compressor Necessary refrigerant quantity Cold power index and the Carnot efficiency Total equivalent warming impact (TEWI) value for the system APPLICATION 2 The linked AC- and engine cooling system shown above was analyzed for one special operating point. The boundary conditions were chosen as: Tubes With the tube simulation model a heat transfer can be considered. The flow of heat between the refrigerant in the tube and the environment can be calculated with a differential equation as follow: 1 dq R = ( tinner tamb ) da 1 1 s + + α α λ i o tube tube with the definitions da tube for the heat transfer area, t inner for the refrigerant temperature and t amb for the ambient temperature of the tube. The heat transfer coefficient α o for the outer side of the tube and α i for the inner side are calculated separately for each step. 3

4 The AC circuit includes an external controlled compressor, a parallel flow type condenser, a standard expansion valve and an evaporator that exists of 4 virtual elements. The simulation target is given as subcooling temperature of 3 [K] (also a fixed refrigerant mass would be an available simulation target). After the simulation run (app. 2 min) the results can be displayed in a graphical post processor. The main results for the above boundary conditions are given as: Refrigerant charge.44 [kg] TEWI 1163 [kg CO 2] Geometric Compressor Capacity 143 [cm 3 ] Compressor Power Consumption [kw] KULI SYSTEM KULI RESISTANCES CFD VELOCITY JOINT RESISTANCES This application is under permanent development to fit the demands of future s and circuit types. Practical usage has shown great accuracy with the reality. KULI-VELOCITY INTEGRATION OF 3-D-CFD VELOCITY DISTRIBUTIONS INTO NETWORK AIRPATH SIMULATION Using one-dimensional flow models based on the network theory is a fast and efficient way to analyze cooling systems. However, this method assumes even air flow distributions within the s of the cooling system. In reality there are uneven air flow patterns in the heat exchangers caused by a fan, by inlets and by non-ideal cooling air channels. In some cases this can have an effect on the heat transfer. Up to now the influence of an uneven air flow could have been considered by dividing the heat exchanger and using different built-in resistances. With KULI-CFD it is possible to model an uneven air flow distribution based on velocity field data. The data for the velocity field usually will be determined by CFD analysis. To model an uneven air flow in KULI the heat exchangers will be combined to a cooling system block. The block has to be divided into a number of elements. For each element a fictive flow resistance will be determined. The flow resistance of an element is defined as a dimensionless characteristic (Zeta value). These Zeta values can be determined in a way that they cause higher velocities in the middle areas and low velocities on the edges. In KULI this system of fictive flow resistances is called Resistance Matrix. APPLICATION 3 To see the advantages and the necessity of the usage of a velocity distribution for a cooling system simulation two comparable cooling systems have been analyzed. The cooling system itself is built by a simple radiator and a fan with a shroud. This means that all the air that is going through the radiator has also to go through the fan. The difference between the two cooling systems is, that in the first case the fan behind the radiator is placed in the centerline and in case two this fan is moved towards one side of the radiator. In the second case the streamlines of the cooling air get a great deflection when going through the system. 4

5 CONCLUSION The two maps show the different velocity fields coming from the CFD analysis. The next two pictures show the coolant temperature distribution for the two cases determined by the uneven air flow through the system. It can be seen that the velocity distribution has in some extreme configurations a great influence on heat transfer and therefor cooling capacity. Often questions appear concerning the calculation time etc. As already described the heat exchangers of the KULI model in case of CFD integration should can be divided into elements. For each of this element the streamline theory is used to solve the energy balancing. Therefor this method often is called 11/2 dimensional method. To avoid big analysis time the following diagrams give some information about the useful resolutions and the related computation time (for a standard 45 MHz PC). CPU-Time [s] Analysis Time No. of Block Elements When looking to the three applications it can be seen that also one dimensional simulation tools bring quite accurate and realistic results. The greatest benefit of such tools are the short analysis time (minutes), the minimum hardware requirements (PC) and the comfortable and easy handling (WINDOWS orientated). For the engineers daily work, the prototyping, optimizing and analyzing it is necessary to provide fast simulation tools that have shown their credibility in practice. To make available methods for an overall simulation of the vehicles heat management is the direction for the future. This paper should show some applications in this field. For the future it is planned to enlarge the software cluster with various modules and methods. REFERENCES [1] T. Sakai, S. Ishiguro, Y. Sudoh, G. Raab, J. Hager: The Optimum Design of Engine Cooling System by Computer Simulation, SAE, International Truck&Bus Meeting 1994, Seattle USA [2] W. Eichlseder, G. Raab, J. Hager, M. Raup: Use of Simulation Tools with Integrated Coolant Flow Analysis for Cooling System Design, VTMS III Conference, SAE Meeting, Indiana, Indianapolis, May 1997 [3] W. Eichlseder, J. Hager, M. Raup: Simulation of cooling systems considering engine heat management, 22nd CIMAC Congress, Copenhagen, Denmark, May 1998 [4] W. Eichlseder: Simulation Techniques in Goods Vehicle Development, FISITA World Automotive Congress, Paris, 1998 T [ C] Coolant Entry Temoperature No. of Block Elements It can be seen that up to a number of 25 elements for a standard car application radiator the CPU-time consumption is in an acceptable range and the results are as accurate as possible. If the resolution it can increase costs for CPU-time without getting better results 5