Pneumatic Nozzle Flapper Control of an. Array of Cantilevers Using a. Single Thermal MicroActuator

Transcription

1 Pneuatic Nozzle Flapper Control of an Array of Cantilevers Using a Single Theral MicroActuator Murad Abu-Khalaf Instructors: Professor Edward S. Kolesar, Texas Christian University Professor Frank L. Lewis, The University of Texas at Arlington ABSTRACT In this work, a Pneuatic Nozzle Flapper is built and siulated. This device is used to stiulate a echanical force that is applied at the sae tie to an array of cantilevers that needs to be pushed echanically. A theral icroactuator functions as a flapper for the pneuatic syste allowing sae aount of force to be applied at the sae tie to all cantilevers. Siulation shows that the idea is feasible and can work at the icro scale. Finite Eleent Analysis is provided for the pneuatic structure and the electrotheral icroactuator. pressure P b is controlled by the nozzle flapper distance x. As the flapper approaches the nozzle, the opposition to the flow of air through the nozzle increases, with the result that nozzle back pressure P b increases. If the nozzle is copletely closed by the flapper, the nozzle back pressure becoes equal to the supply pressure P s. if the flapper is oved away fro the nozzle so that the nozzle-flapper distance is wide, then there is practically no restriction to flow, and the nozzle back pressure P b takes on a iniu value that depends on the nozzle-flapper device geoetry. INTRODUCTION Pneuatic Nozzle Flappers have been eployed successfully in the acro world and been used successfully for several applications. In [5], the pneuatic device was used priarily as a valve in control systes. Unsteady flow force of air jet through nozzle was analyzed. In [1], the pneuatic device was used as a forcebalance pressure control device. In [4], Ogata showed several any applications of the device using it as a PID controller in pneuatic systes. In [3], an extensive study of electrotheral icroactuators, ainly of the inplane tip deflection type. This type of actuators is what we used as a flapper in our design. In figure 1 we show a sketch for a standard Pneuatic Nozzle Flapper device. In operating this device, the flapper is positioned against the nozzle opening. The nozzle back Figure 1. The device can be run in another ode. In this ode you supply a constant flow of fluid at the input, usually air, and by the opening and closing of the flapper, the velocity of the fluid at the output is changed. Using Bernoulli principle, [], we experience a change in pressure at the output too. In this research, we are bringing soe of the successfully eployed ideas in the acro world to the icro world.

2 Microelectroechanical systes fabrication technology (MEMS) allows building such devices in the icro world. The device we are to show in the next section was designed with features that are applicable to surface-icroachining fabrication processes. MUMPS process would suffice for this design. Several software packages were used to siulate the MUMPS process and to ake the design process easier and ore interactive. The ain packages used were MEMSPRO, IntelliSuite. necessary physical properties about the different layers utilized by the MUMPS foundry process. DEVICE THEORY AND DESIGN The layout of the device of interest is shown in figure. This is a scheatic draw using the MEMSPRO software. MEMSPRO software is ainly used to draw the layouts of the device and visualize the in 3D. Also it is able to get cross-sectional views of the specific design like the one we show in figure 3. Figure 3. Device Cross-sectional and 3D views using MEMSPRO Figure. Device Layout using MEMSPRO All the work in MEMSPRO was based on the MUMPS foundry process definition file. IntelliSuite, the other software package used, allows echanical, electrical, fluidic siulation of the devices built in MEMSPRO. By setting the technology files in both software packages to MUMPS, we were able to iport and export the designs easily with IntelliSuite acquiring all the The device diensions are 170 µ 00µ. The electrotheral icroactuator, built as in [3], bends to the left upon applying a voltage signal, and by that closing the nozzle of both pneuatic devices, in other words acting as the flapper. This causes a change in pressure applied to the surfaces of the cantilevers at the output of both pneuatic devices. This pressure change causes the cantilevers to behave echanically the sae way and deflect in a siilar anner. The key point to this is that the devices are syetrical. The pneuatic device is built fro poly1 which is anchored to the substrate. The icroactuator is built also fro poly1. In order to guarantee perfect closure of the nozzle, we included diples at the designated positions on the theral icroactuator bar that will cover the

3 nozzle alost perfectly. The cantilever bea too is built fro poly1, in order for the bea to respond to the output of the pneuatic device, we include diples at the place where the fluid hits the beas. A good start is to begin with the theral icroactuator. The actuator part of the device has been exported fro MEMSPRO to IntelliSuite software for electrotheral finite eleent analysis. Figure 4 shows the response of the actuator for a 10 volts input signal. increased, we did see in siulation that the deflection increases too, for exaple for a 30 volts input, the deflection gets to a value about 0 µ. What usually restrict the usage of higher voltages is the increase in teperature. For the case of 10 volts input, we had the axiu teperature equals C, see figure 5. When the voltage is about 30 volts, the teperature rises to ore than a C. Which eans that we are at the risk of elting the beas before being able to deflect the. The next step in the siulation is to ake finite eleent analysis of the pneuatic device. Before starting the fluidic siulations, we will assue that: 1. When the nozzle is closed, that it is closed copletely without any leakage.. That we have a constant fluid flowing fro the input of the device that coes out copletely, and that the Reynolds nuber of this fluid is below 000. Figure 4. Displaceent response The first assuption allows closing the nozzle with poly1 by anchoring it copletely. And then exporting this odified file to siulate a closed nozzle. The second assuption allows us to do the siulation assuing lainar flow type. And thus ignoring the ixing of fluid layers. Before showing the siulation of the pneuatic device, a short table explaining the ain geoetrical points of the device: Ports Input Output Nozzle Area 8 µ µ 10 µ µ 1 µ µ Figure 5. Teperature distribution The siulations shows that for a 10 volts input, we are able to get a axiu displaceent for poly1 of about 6.5µ at the tip, and about 4.0µ at the center of the bea. This tells us that in order to guarantee a siultaneous closure of the nozzles, we need to place the pneuatic devices close to the top part of the icroactuator, and we need to use a quite long theral icroactuator to do that. If the voltage input is Note that the height of each port is the sae because this is supposed to be the thickness of the first oxide layer which is etched during the fabrication of the device. In order to be able to ake finite eleent analysis for the pneuatic device, we need to get the shape of the internal volue of the pneuatic device. A good way of doing this is to go back to the MEMSPRO layout file, and close all ports of the device by anchoring all of the to the substrate, then aking sure that we don t have etch holes in the process definition file when we build the device the device. This will in turn guarantees that oxide1 will be trapped in the sae volue the air is supposed to flow in. We can then use MEMSPRO to export

4 the oxide1 shape only to the finite eleent analysis packages and work on that only. In the finite eleent siulation package, we then set the boundary conditions on the walls to represents ports as needed. The following figure 6 shows the result in the case when we have the flapper not closing the nozzle. Figure 7. FEA when nozzle is closed In the case when the nozzle was opened, the fluid nored velocity at the output is: Figure 6. FEA when nozzle is opened Opened nozzle Closed nozzle Nored velocity 6e5 icrons/s 8e5 icrons/s The boundary condition is set for 0.01 icroliters/second at the input, while setting the nozzle and the output to be non wall conditions. The nuber 0.01µ L / s is not particularly iportant as we can always increase and decrease this value, but it is a reasonable nuber given the size of the device to start with. Figure 7 on the other hand shows siilar analysis when the nozzle is copletely closed. Looking at the siulation results, we indeed find that the speed of the fluid at the output increases when the nozzle is closed. The pneuatic device design is changed by aking the opening of the nozzle larger, the new diensions of the device is shown in the table below: Ports Input Output Nozzle Area 8 µ µ 10 µ µ 10 µ µ For this new device, the finite eleent analysis shows the following results, see figures 8, 9 for the opened and closed cases.

5 For this new design we get the following change in output velocity Opened nozzle Closed nozzle Nored velocity 4e5 icrons/s 8e5 icrons/s DISCUSSION To find out how uch the pressure difference experienced by the cantilever beas, we use the Bernoulli equation which relates the change in velocity to the change in pressure: 1 = ρ ( V V ) Since the air density is around 1.3 kg/ 3, we have for the first case = 0.17Pa F = A* = e - 01 N For the second case, we have 1 Figure 8. FEA when nozzle is opened = 0.95Pa F = A* = e N Note that the cross sectional area takes into account the area of the diple at the bea s center. Obviously these are very sall forces exerted on the cantilever beas, to increase these forces we need to go back and repeat the siulations for the case when the incoing flowing flow is larger than 0.01µ L / s. To calculate the deflection of the beas we can use the following forula: F( a) y = (3x a), x > a 6EI Fx y = (3a x), x a 6EI Figure 9. FEA when nozzle is closed Figure 10.

6 This equation gives a deflection value of.1e-13 when the applied force is 1.18e-11 N. Since this is a sall value of deflection that will usually go unnoticed, it is iportant to increase the flow at the input until getting a satisfactory pushing force. CONCLUSIONS The siulations shown is done on every part of the whole syste independently. A sort of integrated siulation was not done in which we see all parts working together and influencing the perforance of every other part in the whole syste. This sort of proble liited the ability to find optiu values for the geoetries of the different parts of the device when certain input out relation is desired. But in any case, with reasonable assuptions we were able to separate the different FEA analysis fro each other to ake things sipler. The key value in this device is the input flow, because it is obvious that when we increase this value we will be able to get stronger forces applied to the cantilever beas. Proportional Valve ", Journal Japanese Hydraulic, Pneuatic Society, Vol 6, Issue, p March [] Esposito A., Fluid Mechanics with Applications, Prentice Hall [3] E. Koelsar, S. KO, J. Howard, P. Allen, J. Wilken, N. Boydston, M. Ruff, R. Wilks, In- Plane tip deflection and force achieved with asyetrical polysilicon electrotheral icroactuators, Thin Solid Fils, p , 000. [4] Ogata, K. Modern Control Engineering, Prentice Hall 199 [5] Yaaoto T., Nakada T., Kagawa T., Tanaka K., "Study on Characteristics of Pneuatic Nozzle Flapper Valve Considering Unsteady Flow Force of Jet Through Nozzle", Proc. of ASME/JSME FEDSM 99, F-15, July, 1999 SUMMARY The ain point and ain conclusion of this paper is that we can have a pneuatic switch on the icro scale to turn the power of a fluid on and off, and to use this change in power to ove and push things echanically at the sae tie. The design obviously will be uch ore superior if the siulation was done without any assuption, and if it was possible to run the whole syste as one part. Take for exaple what will happen to the theral icroactuator while closing the nozzle. What about the resistance it faces, is the force of the actuator enough to overcoe such resistance. Another proble is how the heat of the actuator will affect the fluid flow inside the pneuatic device. These are all iportant points that where not considered in this research. REFERENCES [1] Araki K., Chen N., Ishino Y., "Characteristics of a Force-Balance Nozzle- Flapper Type Pneuatic Pressure Control