Virtual,Thermal,Test,System,Based,on,Simulink,and,Comsol,Co-Simulation

ZHOU Xingguang,LIU Bo,JIANG Gang,WANG Zhaoxi,LIANG Di,LIU Shiling

Shanghai Spaceflight Precision Machinery Institute,Shanghai 201600

Abstract:A virtual thermal test system was built through the co-simulation using Simulink and Comsol to realize the complete virtualization of the thermal test.Using the co-simulation technology,comprehensive simulation analysis of the control system,electric field and thermal field was realized.The data state of each observation point could be directly observed at one time,including the output state information of the power amplifier,the output state information of the heater,and the thermal state information of the test unit.The virtual thermal test system has a predictive and guiding role for engineering thermal tests,and can realize thermal environment simulation beyond the existing thermal environment ground simulation capabilities,providing a basis for the development of future models.

Key words:Simulink,Comsol,co-simulation,virtual thermal test system

With the significant increase of the design value for aircraft flight speed,the high-temperature thermal environment caused by aerodynamic heat becomes more and more severe[1-5].Aerodynamic heating will reduce the stiffness and strength of the aircraft structure,and produce thermal stress,thermal strain and material ablation,which will increase the internal temperature and worsen the working environment in the cabin[6-11].Thermal test is an effective method to study the problem of pneumatic heating.It has become an important means of structural design,strength and reliability analysis,product performance inspection and appraisal.To date,the development of engineering thermal testing has encountered bottlenecks including high heat flow output heater design,accurate tracking of high dynamic spectral lines and thermal test at high heat flux.

Virtual thermal testing based on software engineering simulation can greatly improve the technical level of ground simulation of the thermal environment,break through the above bottlenecks in thermal testing.It can predict and guide the engineering thermal test,and realize a thermal environment simulation beyond that of the ground simulation ability of the thermal environment.It provides a basis for model development,so as to speed up the progress of model development,improve the quality of model products and reduce the cost of model development.

Due to the complexity of the thermal test,it is very difficult to establish a virtual thermal test.At present,there is no comprehensive research in this area in China.Reference [12]analyzed how to carry out a modular virtual thermal test from the perspective of the composition of the thermal test system,but there was no specific implementation method identified.Reference [13] realized the model establishment of the tested product and heater.Reference [14] presented a virtual thermal test method for quartz lamps based on MSC.Nastran/Patran used finite element analysis technology to realize a thermal test virtualization.Looking at the development of domestic virtual thermal tests,the public literature mainly focuses on the virtualization of heaters and tested products.In addition to the heater and the tested product,the thermal test system includes a heating power supply and a control system.In the public literature,there is no relevant information on realizing a virtual thermal test method for the virtualization of the entire thermal test system.

In order to realize the virtualization of the entire thermal test system,the virtual thermal test method shown in Figure 1 was established.Simulink was used to realize the virtualization of the control system,Comsol for the virtualization of the thermal field and electric field,and the data interaction between the control system and electric and thermal fields.

Figure 1 Overall block diagram of virtual thermal test

The input signal of the module is the signal of the load spectrum,the output of the module is the conduction angle corresponding to the power conversion device.During the control process,the required control parameters such as proportional,integral and differential can be freely set before the virtual thermal test is implemented.The power conversion module provides raw energy with AC power.The power conversion part can be an AC/AC converter or an AC/DC converter constructed through power electronic devices such as thyristor,IGBT,and MOSFET.The conduction control of the power electronic device is controlled by the output of the control module.Depending on the nature of a given load spectrum during a virtual thermal test,the signal used for feedback is chosen to be heat flow information,or temperature information.The Simulink-based control model is shown in Figure 2.

Figure 2 Simulink-based control model

During the co-simulation process,when the Simlink-based control model is run,the Comsol-based electric field and thermal field models will run automatically.At any observation point in Simulink,by adding an oscilloscope observation module,the signal changes at each observation point during the virtual thermal test can be observed.

The overall design idea based on the co-simulation of Simulink and Comsol was to analyze the thermal characteristics of a tungsten filament heat source through the Comsol software,realize the virtualization of the thermal field and the electric field,and convert the input electrical signal into a thermal signal.The control of thermal signals was realized through Simulink.The overall design block diagram is shown in Figure 3.

Figure 3 Overall model design block diagram

In Figure 3,the difference between the given signal and the output heat flow is sent to the PID controller,the signal output by the controller is processed and transformed into an angle signal,and a pulse generator was used to generate two pulse signals with a phase difference of 180˚,which were transmitted to the phase shifting transformer,and then converted a voltage signal into power signal,transmitting it to the Comsol module,where the heat flow signal is calculated to form a closed-loop control system.

The Simulink closed-loop control system model is shown in Figure 4.The model includes the given load spectrum information (heat flow),AC voltage conversion device,closed-loop control module,and feedback heat flow signal.

The PID controller is used for control,mainly using PI control.The output signal of the PID controller is converted into an angle signal through a series of mathematical calculations and transmitted to the pulse generator,and the first and fourth pulse signals of the pulse generator are transmitted to the phase-shifting transformer where the phase difference between the two signals is 180˚.

The AC voltage regulator mostly realizes the control of the power supply by means of phase-shift trigger voltage regulation.The phase-shift trigger voltage regulation is to adjust the conduction angle by controlling the phase of the trigger pulse in the positive or negative half cycle of the alternating current,so as to achieve the purpose of changing the average value of the output voltage.By controlling the turn-on phase of the thyristor in each half cycle,the effective value of the output voltage can be easily adjusted.The quartz lamp can be approximately regarded as a pure resistive load.The schematic diagram of the AC voltage regulation circuit under resistive load is shown in Figure 5.The thyristor shown in the figure can also be replaced by a triac.In the positive half cycle and negative half cycle of the AC power supply,the output voltage can be adjusted by controlling the opening anglesαof VT1and VT2respectively.The positive and negative half cycle whereαis the start time (α=0) the voltage zero-crossing time.In steady-state conditions,the positive and negative half cycles should be made equal.It can be seen from Figure 5 that the waveform of the load voltage is a part of the waveform of the power supply voltage.The waveform of the load current (i.e,the power supply current) and the load voltage are the same.Therefore,the output voltage can be controlled by controlling the trigger delay angleα.

Figure 4 Simulink closed-loop control system model

Figure 5 Resistive load single-phase AC voltage regulator circuit

After the voltage is converted into an effective value,it is processed as a power signal and transmitted to the Comsol module.At this time,the background server will automatically call on Comsol to perform an operation and output the temperature and heat flow signals.The difference between the heat flow signal and the given heat flow signal is transmitted to the PID controller to complete the entire closed-loop control.

Using the established model,not only the characteristic analysis of the load spectrum can be tracked,but also the characteristic analysis of the heater can be obtained.Data analysis is mainly carried out through the following aspects.

4.1 The Characteristic Analysis of the Load Spectrum Tracking

For the model established above,when the simulation step is 20 ms,the output of the PID controller is shown in Figure 6.The RMS voltage in the model is shown in Figure 7.The PID control tracking is shown in Figure 8.

Analysis of Figure 8 reveals that the control effect before 3 s is poor,which is mainly reflected in the large hysteresis due to the inertia of the thermal characteristics of the quartz lamp within 1 s,which is purely a forward error.In the 2−3 s time period,because the output of the control quantity is too large,a large overshoot occurs.

Figure 6 Output of the PID controller

Figure 7 Voltage RMS change

Figure 8 Heat flow tracking performance

The tracking effect in the 0−5 s time period is similar to that in the engineering thermal test process,which shows that the established model is reasonable.

Using PID in the model,the stable tracking of the heat flow signal can be realized.

4.2 The Characteristic Analysis of the Heater

During the thermal test,the thermal inertia is mainly caused by the tungsten filament of the quartz lamp.Using the established model,with a fixed input power,the heat flow stabilization time and stable value corresponding to different tungsten wire diameters are shown in Table 1.As shown in Table 1,the smaller the diameter of the tungsten wire,the shorter the heat flow stabilization time and the smaller the thermal inertia,which is more conducive to the heat flow control.

Table 1 Heat flow response characteristics corresponding to different tungsten wire diameters

The quartz lamp heater relies on thermal radiation for energy transfer,and the radiation energy is related to the temperature of the tungsten filament.With the help of the model,the heater characteristics under different input voltages were analyzed,as shown in Table 2.Comparing with Table 2,it was found that the influence of air on the radiation transfer of heat flow is small and can be ignored.

Table 2 Analysis of air-to-heat transfer characteristics

Through the co-simulation technology of Simulink and Comsol,a virtual thermal test system was built,which realized the complete virtualization of the thermal test;co-simulation technology was used to realize the comprehensive simulation analysis of the control system,electric field and thermal field;The monitoring of the output state data of the device,the thermal state of the test unit and other data had a predictive and guiding role for engineering thermal tests,and could also achieve thermal environment simulation beyond the existing thermal environment ground simulation capabilities.

推荐访问:Test System Virtual