Ultrasonic Flow Sensor - Computer Action Team



Maseeh College of Engineering and Computer ScienceElectrical Engineering DepartmentMechanical Engineering DepartmentSenior Capstone Final Report – 2010Group Members:Tyler JonesShi YangJames MillerAaron PooleAcademic Advisor:Dr. James MorrisIndustry Advisor:Eric Bond, Project Engineer1 - Executive SummaryEdwards Vacuum pumps develop thermal abatement systems that assist in silicon wafer production, among many other products. These systems are essential for handling exhaust gases so that they are cooled and any acid larger particulates are scrubbed out of the system.It essential that all exhaust gas makes it way to the combustor where it is burnt up and then what is left is sent to the scrubber. Failure to do so, could affect the efficiency of the process as well as well limiting the life of the quadrant pipe and other components prior the quadrant. Since most gases flowing through the system can be very corrosive (Cl2 and F2). Knowing the volumetric flow rate would allow engineers and technicians to be able to detect a blockage as well as being able to analyze process efficiency.The project sponsor, Edwards Vacuum pumps, would like a volumetric sensor that can detect flow of corrosive gases such fluorine and chlorine. The sensor should be able to handle high temperatures, and it is key that the gas flow is not obstructed. This project would then be further developed by the R&D department over the next 3–4 years.The design team has developed a custom ultrasonic sensor that is built in to the existing quadrant to measure volumetric flow of the gas. This sensor operates by ultrasonic transducers sending and receiving ultrasonic pulses that the circuit uses to calculate the flow rate based on the difference in transit time of the pulses. Although built-in ultrasonic sensors exist, none can operate well at these high temperatures. This is accomplished by using custom-designed transducer, as well as using an insulating sleeve to hold and house the transducers.Table of Contents TOC \o "1-3" \h \z \u 1 - Executive Summary PAGEREF _Toc263665953 \h 22 - Mission Statement PAGEREF _Toc263665954 \h 63 - Introduction and Background Information PAGEREF _Toc263665955 \h 64 - Product Design Specifications PAGEREF _Toc263665956 \h 85 - Top Level Design Considerations PAGEREF _Toc263665957 \h 105.1 - Thermal Technology PAGEREF _Toc263665958 \h 105.2 - Coriolis Technology PAGEREF _Toc263665959 \h 105.3 - Differential Pressure Technology PAGEREF _Toc263665960 \h 115.4 - Ultrasonic Technology PAGEREF _Toc263665961 \h 125.5 - Technology Selection Matrix and Final Design Consideration PAGEREF _Toc263665962 \h 136 - Electrical Design PAGEREF _Toc263665963 \h 146.1 - Research and Overview PAGEREF _Toc263665964 \h 146.1.1 - Research Findings and Interpretation PAGEREF _Toc263665965 \h 146.1.2 - Overview of the Design PAGEREF _Toc263665966 \h 146.2 - Electrical Circuit Design PAGEREF _Toc263665967 \h 156.2.1 - Level-0 Block Diagram PAGEREF _Toc263665968 \h 156.2.2 - Level-1 Block Diagram PAGEREF _Toc263665969 \h 166.2.3 - ATmega325P Microcontroller PAGEREF _Toc263665970 \h 176.2.4 - TDC-GP2 Ultrasonic Flow Sensor Chip PAGEREF _Toc263665971 \h 176.2.5 - Switching Network PAGEREF _Toc263665972 \h 186.2.6 - FIRE Pulse Amplification PAGEREF _Toc263665973 \h 196.2.7 - Receiving Circuitry PAGEREF _Toc263665974 \h 196.2.8 - 7-segment Display PAGEREF _Toc263665975 \h 206.2.9 - Programming PAGEREF _Toc263665976 \h 217 - Mechanical Design PAGEREF _Toc263665977 \h 247.1 - Research PAGEREF _Toc263665978 \h 247.1.1 - Research Findings and Interpretation PAGEREF _Toc263665979 \h 247.1.2 – Standards PAGEREF _Toc263665980 \h 247.1.3 - Overview of the Design PAGEREF _Toc263665981 \h 257.2 - Design PAGEREF _Toc263665982 \h 257.2.1 - Drawings of the Quadrant PAGEREF _Toc263665983 \h 257.2.2 - Design of the Modified Quadrant PAGEREF _Toc263665984 \h 257.2.3 - Weld Stub Fitting PAGEREF _Toc263665985 \h 267.2.4 - Transducer Press-Fit Sleeve PAGEREF _Toc263665986 \h 277.2.5 - Transducers PAGEREF _Toc263665987 \h 278 - Final Design PAGEREF _Toc263665988 \h 288.1 - Complete Design PAGEREF _Toc263665989 \h 288.2 - Sensor Housing Assembly PAGEREF _Toc263665990 \h 288.3 - Insulating Transducer Sleeve and Modified Long Neck Flange PAGEREF _Toc263665991 \h 308.4 - Electrical Housing PAGEREF _Toc263665992 \h 308.5 - Main Circuit Schematic PAGEREF _Toc263665993 \h 328.6 - 7-Segment Circuit Schematic PAGEREF _Toc263665994 \h 329 - Testing and Evaluation PAGEREF _Toc263665995 \h 339.1 - Electrical PAGEREF _Toc263665996 \h 339.2 - Proving the Accuracy PAGEREF _Toc263665997 \h 359.2.1 - Rotameter Comparative Testing PAGEREF _Toc263665998 \h 359.2.2 - Differential Pressure Readings PAGEREF _Toc263665999 \h 369.2.3 - Results PAGEREF _Toc263666000 \h 3610 - Future Considerations PAGEREF _Toc263666001 \h 3710.1 - Electrical PAGEREF _Toc263666002 \h 3710.2 - Mechanical PAGEREF _Toc263666003 \h 3810.3 - Transducers PAGEREF _Toc263666004 \h 3811 - Conclusion PAGEREF _Toc263666005 \h 3912 - References PAGEREF _Toc263666006 \h 4013 - Appendix PAGEREF _Toc263666007 \h 40Appendix A - Bill of Materials PAGEREF _Toc263666008 \h 40Appendix B - Manufacturing Instructions PAGEREF _Toc263666009 \h 43Appendix 1B - Original Quadrant Drawing from Edwards PAGEREF _Toc263666010 \h 43Appendix 2B - Modifications to the Quadrant PAGEREF _Toc263666011 \h 44Appendix 3B - Transducer Fittings PAGEREF _Toc263666012 \h 45Appendix 4B - Electrical Housing Box PAGEREF _Toc263666013 \h 46Appendix C - The Program PAGEREF _Toc263666014 \h 47Appendix D - Calculations PAGEREF _Toc263666015 \h 53Time of Flight PAGEREF _Toc263666016 \h 53Volumetric Flow Rate PAGEREF _Toc263666017 \h 54Reflection and Transmission Coefficients PAGEREF _Toc263666018 \h 54Flow Analysis PAGEREF _Toc263666019 \h 55Heat Transfer Analysis PAGEREF _Toc263666020 \h 57Mechanical Analysis PAGEREF _Toc263666021 \h 60Appendix E - Top Level Research PAGEREF _Toc263666022 \h 65Thermal Sensing Technology PAGEREF _Toc263666023 \h 65Appendix F - Manufacturing Drawings PAGEREF _Toc263666024 \h 75Appendix G - Raw Data PAGEREF _Toc263666025 \h 762 - Mission StatementTo develop a volumetric flow sensor that measures the volume of gas flowing through a quadrant that feeds gas into the combustion chamber of the thermal abatement system.The goal of this project is to employ the time-of-flight method of ultrasonic pulses to measure the volumetric flow of gas in a 1-inch stainless steel pipe. The time difference between an upstream and downstream pulse can be calculated, and from that time difference and the cross-sectional area of the pipe the volume per second can be calculated. We will not be designing for immediate application, but rather to prove that the technology will work for the particular setup for which Edwards wants to employ it.3 - Introduction and Background InformationThe purpose of a thermal abatement system is to cool exiting gas and remove harmful particulates that can be hazardous or harmful to the environment. Some of these exiting gasses consist of fluorine (F2) and chlorine (Cl2). Figure 3-1 shows a schematic that describes the process in which gas enters from the top of the combustor. With the addition of heat, these gases can be made into combusting compounds (CO2, H2O, H2). [1]Figure 3.1 - Typical Edwards’ thermal abatement system that uses an inward-fired combustor. [1]The exiting gas leaves the combustor, is cooled, and then is released to the atmosphere. Any acids or hard particulates are subsequently scrubbed out of the system. However, to ensure that this process works, the gas needs to be controlled in such a way that the velocity, the amount of gas, and quality of gas is in a form that can be combusted before it reaches the combustor. Figure 3.2 shows this part of the process. Process gas enters the quadrant (pipe) into which an auxiliary pipe injects oxygen to the system. A scrubber with injected methane then helps clean the gas. In this process, it is essential to have an accurate reading of the flow rate of the gas into the combustor. This knowledge ensures that the proper combusting compounds are produced in the Combustor and also notifies someone when a clog that restricts flow rate occurs, which can induce pressure and increase corrosion. [1]Figure 3.2 - Cross-section of the quadrant and scrubber system located on top of the combustor. [1]The project proposed by Edwards to the design team was to design and build a volumetric flow sensor capable of detecting process gas flows that may consist of fluorine and chlorine gas. The design team’s tasks were to research existing technology and create a custom solution that can be implemented in Edwards’ thermal abatement systems.4 - Product Design SpecificationsTable 4.1 - Overview of the design constraintsGeneral Requirements:Budget must stay within specified limits.The project has a budget of $5,000 for production of at least one sensor capable of sensing flows within one of the quadrant tubes. Further funding will not be provided but the ability to sense multiple tubes would be optimal.Accuracy must be close to what is available in an off the shelf flow meter.The accuracy of the flow meter must be comparable to similar models available on the market. Typical accuracies are 5% error or less. Environmental Requirements:The flow sensor needs to survive a corrosive environment.The flow sensor must withstand corrosion from chlorine and fluorine gases as well as 304 stainless steel; which is what the quadrant tubes are made from. Typical lifetime of the quadrant tube is 7 years with 12 month service intervals. The sensor element must handle high temperatures.The temperature of the stainless steel tubing can reach 180 °C at the upper o-ring sealing surface. The gas flow temperature is commonly less than 35 °C.The sensor should not impede flow or cause buildup of particulate from gas flow.The gas flow contains particulate that can build up on protruding surfaces. Pressure drop resulting from sensor components will need to be near zero or negligible.Electrical Requirements:The entire range of possible flows will need to be measured.Flow can range from 10 up to 100 SLM requiring the sensor to have a turndown ratio of 10:1. Flowrate will be calculated with a moving average and output at a 1Hz frequency for data logging.Power for the sensor must come from available power supply currently in cabinet.The power available in the cabinet is 110VAC and 24VDC. Power supplied directly to the circuitry can be regulated to 12V, 5V and 3.3V DC within the designed circuitry.Output will need to be logged in a computer for analysis.Currently in the cabinet is a PLC that can accept a 0-5V DC signal proportionate to the flowrate or a digital RS232 input.5 - Top Level Design ConsiderationsThere are four types of flow sensing technologies which we thoroughly researched for design considerations. They included thermal, coriolis, differential pressure and ultrasonic sensing technologies.5.1 - Thermal TechnologyThermal sensors work on the principle of heat transfer. A heated element with a known temperature is inserted into a pipe that contains a gas flow. As the gas passes the heated element, heat is taken away from the heated element and a temperature difference occurs. An RTD (resistance temperature detector) senses the temperature change and changes its resistance, which then change the voltage read-out. This change in temperature can be directly correlated to the amount of flow passing through the pipe. [2]Figure 5.1 - A thermal insert able flow sensor from Serria Instruments.The advantages of using this technology are that it is simple and can be easily built with the appropriate components. Some modification to the pipe is required, but it is minor. However, a flow restriction is created, and a pressure drop occurs. Most manufacturers of thermal sensors claim the pressure drop is low, although experimental data were not available to confirm how low the pressure drop was. Another disadvantage is that gas properties must be known in order to calculate a “mass flow”, which then can be used to calculate the flow rate. Because the gas is mixed and is not always known, this was not considered a viable option.5.2 - Coriolis TechnologyCoriolis flow sensors involve a bent tube or series of tubes through which the fluid travels . These tubes are then excited by a uniform forcing pulse that causes the pipe to slightly oscillate. Sensitive sensors pick up the motion of the oscillating pipe, one before the forcing exciter and one after. As the fluid moves through the pipe, the pipe begins to “twist” and causes a slight change of phase between the two sensors. The phase change is proportional to the mass flow of the gas, from which the volumetric flow rate can be derived. Furthermore, the sensor is also able to measure the density of the fluid flowing through the pipe. For this type of sensor, it is possible to measure the volumetric flow without knowing the characteristics of the gas. Typical commercial coriolis flow sensors have exhibited accuracy ±0.5%, which exceeds expectation. Figure 5.2 - A coriolis flow sensor constructed in a delta configuration.Although coriolis flow sensing technology shows great promise for accuracy as well as for an emerging market, the technology has its drawbacks. One disadvantage is that it measures mass flow versus direct fluid flow. Although the density of the gas can be approximated, this would affect accuracy of readings as the gas will change density with different temperatures. Another concern is sensitivity. Due to the large amount of ambient vibration, misreading from the vibration transducers may occur. There also appears to be no evidence that this technology will work in a vertical configuration. The last major issue to overcome, and possibly the most important, is the amount of pipe required to implement a coriolis sensor. Most researched existing sensors use more space and pipe than would be permissible considering usable space.With all of these consequences, the design team decided that this choice of technology would not be the best option.5.3 - Differential Pressure TechnologyDifferential pressure technology is the simplest of the sensing technologies because it uses an obstruction to cause a pressure drop so that pressure can be sensed before and after the pressure sensor. This is based on the Bernoulli equation, in which the velocity can be found, leading to finding the flow rate, by knowing the pipe and obstruction geometry. Because of the design specification that requires no flow restriction; differential pressure technology cannot be selected as a design solution for the flow sensor. Although gradual restrictions exist, such as a venturri tube, pressure drops still occur and can enhance corrosion.Figure 5.3 - A typical example of differential pressure drop being used for flow measurement.5.4 - Ultrasonic TechnologyUltrasonic sensing relies on transducers sending ultrasonic pulses back and forth along different sections of the pipe. An ultrasonic pulse will take longer to travel a particular distance against a gas flow than it would with a gas flow. Using this method, known as the time-of-flight method, the difference in the transit times can be measured. With knowledge of the pipe’s geometry, taken together with the difference in the transit times, the flow rate can be calculated. Both clamp-on and integrated sensors exist, thus flow does not necessarily have to be obstructed. This option is ideal because it allows for no, or very little, modification to the existing pipe in order for it to calculate the flow rate.Figure 5.4 - Typical example employing the time-of-flight method.One disadvantage is the availability and cost of sensors used in such an environment. Another disadvantage is the diameter of the pipe and the travel distance associated with a smaller pipe.5.5 - Technology Selection Matrix and Final Design ConsiderationOnce we had several possible technologies to consider, we needed to determine how to decide which would work best for implementation. To do this, we created a design matrix that would rate the technology based on six key design parameters. We gave each technology one of three ratings: low means that there is a little concern regarding the parameter, medium means that there is some concern regarding the parameter, and high means that there is a lot of concern regarding the parameter.The differential pressure method was not considered due to the significant pressure drop associated with it. As can be seen from the selection matrix, the choice with the fewest design issues and concerns is the ultrasonic sensing technology. Based on this, the team selected this technology for further development as the team’s sensor.With approval of the concept design from the team’s industry advisor, further research was conducted on specific components needed for construction of the ultrasonic flow sensor. One such component was the ultrasonic transducers. The initial plan was to purchase General Electric Panametrics transducers, which were designed for measuring gas flow. However, after further research, we discovered that these transducers were not available for purchase individually and could only be purchased by buying the complementary electronics, which was not an option. During discussion with engineers from General Electric, we discovered that a clamp-on ultrasonic flow sensor would not be optimal. To understand why, we decided to calculate the reflection and transmission coefficients of sound. The calculations showed that 99.996% of the ultrasonic wave would be reflected at the boundary between steel and gas because the density change at the medium is too great. Thus, not only would we have nearly 100% of the transmitted wave’s initial intensity reflected at the steel-to-gas medium but also nearly 100% of the transmitted wave would be reflected at the gas-to-steel medium when the wave is received.The team still considered ultrasonic sensing to be the best option, but now it would have to be built into the existing pipe. This was not preferred by Edwards; however, minimal pipe modification would be required.6 - Electrical Design6.1 - Research and OverviewOnce the ultrasonic time-of-flight method was chosen, research was required in order to understand the technology before any design work could be done. There were several key areas of research, such as the ultrasonic transducers themselves, the circuitry involved with the transducers, the control of the pulses and any noise associated with the outside environment, and efficiently calculating the flow rate.6.1.1 - Research Findings and InterpretationThe transducers require high voltage levels in order to adequately create an ultrasonic pulse capable of being detected. Typical voltage levels are 100 to 500VDC. Because the transducers we chose for this project use a ceramic oscillator, the amplitude of the ultrasonic pulse produced by the transducer is proportionate to the amplitude in the voltage difference across the ceramic element. In order to produce the high voltage, a 40:1 DC pulse transformer would be used along with a FET transistor.In order to control the upstream and downstream pulses and to keep the amount of circuitry required to a minimum, we are using two analog components that contain three independent switches. These would be controlled by the microcontroller. The number of switches is necessary because of the high voltages in the circuit. The switches are also used to reduce the noise in the system by opening for short periods of time to allow a pulse to pass through.Finally, we needed to have a circuit to receive the ultrasonic pulse from the receiving transducer. This receiving circuit would do two things: it would first adjust a high voltage down to .7 V with a diode, and second, it would filter and amplify the signal to 5 V. Once the receiving pulse has been filtered and amplified, it becomes the STOP pulse, which is routed to the TDC-GP2 chip.6.1.2 - Overview of the DesignThe electronics are broken into two parts. The first is the main PCB, which houses the inputs and outputs, microcontroller, TDC-GP2 chip, switching network, and the amplifier/receiving circuits. The second is the 7-segment display PCB, which houses the 7-segment BCD decoder and the 7-segment displays. The two PCBs are connected by a 12-pin connector. Cost constraints limit the size of the PCBs to 10 square inches and only two layers.6.2 - Electrical Circuit Design6.2.1 - Level-0 Block DiagramModuleUltrasonic Volumetric Flow Sensor Inputs- DC signal from T/R Transducers - Power: 12 VDC regulated to 5 VDC Output ? VDC signal to computer systemFunctionalityThe sensor will use the travel time of an up-stream and down-stream ultrasonic pulse. The travel time will then be used by a microcontroller to calculate the volumetric flow rate. The calculated rate will be sent to a computer system. 6.2.2 - Level-1 Block Diagram6.2.3 - ATmega325P MicrocontrollerThe Microcontroller (MC) synchronizes the interactions between itself and the TDC-GP2 sensor chip and the direction of the up and down stream pulses. It is programmed through its JTAG interface and communicates with the TDC-GP2 through its SPI serial interface. The SPI serial interface is also used to program the TDC-GP2 chip.Pins PC0 through PC3 are used to control the switching network, which routes the up and down stream pulses. Pins PG0 through PG4 control which of the 7-segment displays is on during a refresh. Refreshing occurs 50,000 times per second. Pins PA0 through PA3 are used to send the BCD flow rate to the 7-segment BCD decoder on the 7-segment circuit.It operates at a core clock frequency is 16MHz, AVR 8-bit instruction with 32kB of programmable flash memory. The core voltage runs at 5V and has 54 programmable I/O lines. Its peak operating temperature is 85°C which does not meet the 125°C specifications; however, the electronics are stored in another portion of the cabinet which houses the quadrants. The temperature where the electronics are stored is estimated to never exceed 60°C.6.2.4 - TDC-GP2 Ultrasonic Flow Sensor ChipThis chip is specially designed for ultrasonic flow measurement employing the time-of-flight method. Another feature is its temperature measurement, but we are not using that in this project. The chip has a core clock frequency of 4MHz and optimizes a 50ps bus to keep track of the time of flight. It operates at a core voltage of 3.3V and has an I/O voltage of 5V. The FIRE pulse that is produced during time-of-flight measurement is 5V and 48mA.In order to optimize the performance of this chip, the amount of travel time is needed. There are two different measurement ranges: the first has a measurement range of 3.5ns to 1.8?us, and the second has a measurement range of 500ns to 4ms. For our purposes, we calculated that it would take roughly 400us for the ultrasonic wave to travel through the pipe. Thus, measurement range two was selected (time delays in the circuitry were ignored).Measurement range two has one channel with a typical resolution of 50ps and is able to read a rising and/or falling edge. The bus is initialized before the FIRE pulse is sent and begins counting when the FIRE pulse is sent (the two pins are connected).6.2.5 - Switching NetworkThe switching network is used to route the upstream and downstream pulses. It is directly controlled by the MC and is synchronized with the firing sequence of the TDC-GP2 chip. There are five switches; two switches route the FIRE pulse from the TDC-GP2 to the upstream or down stream amplifier circuit, two switches route the received pulse, and the lastly there is a switch that routes the amplified received pulse to the STOP pin on the TDC-GP2.We are using two SPDT-MAX4619CPE analog switch ICs. Each IC contains three switches, operates at 5V, and has an on time of 15ns. Figure 6.1 shows how the switches are connected to the amplifying and receiving circuits.Diodes are used to lower the high-voltage pulses down to .7V so that the sensitive IC is not overloaded during operation. When a downstream pulse is sent, the receiving circuit attached to the transducer is disconnected by the switch. Likewise, when the downstream pulse is sent, the amplifier circuitry on the receiving end is turned off by a FET transistor. Refer to Figure 6.1.6.2.6 - FIRE Pulse AmplificationFigure 6.1 - DC Pulse Transformer.404177573660In order to optimize the performance of the transducers, the FIRE pulse from the TDC-GP2 needs to be amplified from 5V to over 100V. The current consumed by the transducer is negligible. The transducers that we selected have a maximum driving voltage of 400Vp-p.Several attempts to use an analog switching amplifier failed to produce the desired square wave amplification that we desired. However, there is another option that will create a short high-voltage burst. The amplification in our circuit is done by a 1:40 DC pulse transformer. With a supply voltage of 5V, we are able to produce a 200Vp-p spike that is used to drive the transducer.The pulse is then passed through a rectifier circuit to try to square off the pulse as much as possible before it is passed to the transducer. A 1? watt 1k ohm resistor absorbs any excess current conducted by the diodes.6.2.7 - Receiving CircuitryThe receiving circuitry consists of an amplifier with a maximum DC amplification of 100. The circuit will be used to amplify the voltage of a received pulse to a maximum of 5V. Once the pulse is received by the transducer, it will immediately be passed through a high-power resistor to absorb any excess current. The pulse is then lowered from whatever voltage it is received at to the .7V of the forward biased diode. Some filtering is done through the 4.7nF and 100pF capacitors, and the circuit is capable of being used as a high- or low-pass filter (the capacitor in the feedback loop).Figure 6.2 - Switching, Amplifier, and Receiving circuits are shown.42964101784356.2.8 - 7-segment DisplayFigure 6.3 - 7-Segment DisplayFlow rate information is received from the MC in the form of a BCD. A 7-segment common anode decoder converts the BCD and displays the value on the 7-segment LED (LED). A latch system is used to control which LED is being refreshed at any given time. The MC is used to turn on and off each of the LED’s at a rate of 50,000 times per second. Five blue Kingbright LED’s are used to give us a maximum flow reading of 999.99 slm (standard liters per minute). 6.2.9 - ProgrammingIn sections 6.2.3 through 6.2.8 there were six parts of the total circuit described. A program is needed to control when the TDC sends a pulse, which direction the pulse is to be sent, calculate the average flow rate from a series of time differences from the TDC, and display that flow rate on a 7-segment display. This section will break down these different parts and briefly show how it is done. The full length program can be found in the appendix on page 47.First, the MC controls the TDC. Their communication is over their built in SPI interface. The MC sends an operation code to the TDC, and the TDC recognizes the code and does the operation. The operations are up or down stream FIRE commands.Figure 6.4 - Operation code summary.In the table 6.4, ADR2, ADR1, and ADR0 are the configuration registers within the TDC chip. In order to communicate with the TDC, these registers need to be configured. Writing operation: If we wanted to configure register one of ADR, the 0x81 operation code is sent. Then a 24bit register configure value is sent. During this time, it is especially important to not call /SS between those two steps. Reading operation: If we wanted to read the value of register one, the 0xB1 operation code is sent to the TDC chip and the 24bits of information would be transmitted to the MC. After the operation code is sent, it is especially important that you do not stop CLK or call /SS. There are other operation codes which can be sent to the TDC chip. Some examples are: 0x70 which initializes the TDC, 0x50 which resets the TDC, and 0x01 which will start a firing command. All three of these commands are used for the basic functionality of the program. Below is a code segment of how these are used.REG0=0X80AB5668; REG1=0X81214200;REG2=0X82E03200;REG3=0X83200000;REG4=0X84203400;REG5=0X82000000;In the program, the TDC’s operational workflow is as follows:SPI-send Reset 0x50 SPI-send REGCONFIG 0x8X+24bit SPI-send Initialize 0x70 SPI-send start firing 0x01 SPI-read read state 0x B5SPI-send fire 0x01;There are some specific things to pay attention to within that sequence. Because of how the SPI functions, data is transmitted and received simultaneously. Therefore, if we want to read data from the TDC, /SS must not be called and a 0 must be written to the buffer in order to keep CLK from stopping during transmission. Below is a code segment which shows how the SPI sends and reads data during an operation: if(bitlength==32) { for(int i=0;i<4;i++) { datat>>=(24-(i<<3)); SPDR=datat; while(!(SPSR&(1<<SPIF))); temp[i]=SPDR; datat=data; }Secondly, we use ports on the MC to directly control the switching, and a counter is used to keep track of the next direction that the FIRE pulse is to be sent. If the counter is an odd number, the switch is in the up-stream direction, and for an even number it is in the down-stream. Below is a code segment showing the switching direction:if(direction==0) // updirection direction=0{ PORTC=(1<<PORTC0)|(0<<PORTC1)|(0<<PORTC2)|(1<<PORTC3);}if(direction==1){ PORTC=(/*0*/0<<PORTC0)|(1<<PORTC1)|(1<<PORTC2)|(0<<PORTC3);)Third, the MC controls the 7-segment LED display circuit. The LED circuit is composed of two parts, a decoder (74ls47) and five 7-segments. Each 7-segment is Anode connected to one port of the MC. The decoder reduces many steps of the data transformation because the MC can simply output a BCD value to the ports which are connected to the decoder. The BCD is then converted to display the value on one 7-segment at a time.In order to display the flow rate properly, we want to turn on each of the 7 segments in a sequence. This is done by an internal timer on the MC to control the rate at which the 7-segments are turned on or off. The rate should be at least 60Hz so that the human eye cannot see when a 7-segment is on or off. During this sequence, only one of the five 7-segments is on. The timer used is the 0 timer. Before the MC can send the information to the decoder, the flow rate needs to be converted to BCD. The maximum flow rate that can be displayed is 999.99 slm and a simple algorithm is used to send the information. data= flowrate*100+0.5; simplify the fraction part numberled_buf[4] = (data / 10000); store the highest bit to led display bufferNUM1 = data % 10000; calculate the number leftled_buf[3] = (NUM1 / 1000); NUM1 = NUM1 % 1000;led_buf[2] = (NUM1 / 100);NUM1=NUM1%100;led_buf[1] = (NUM1 / 10);led_buf[0] = (NUM1 % 10);A code segment of how the timer is initialized:DDRG|=0x1f;DDRA|=0x0f;TCCR0A=0b00111101; //CTC,1024,Set OC0AOCR0A=0x0f; //top valueTIMSK0=0b110; //OCIE0AAt the same time the flow rate is being displayed on the 7-segment display, it is being transferred over UART to an external computer system. Due to the fact that the circuit is not 100% complete, data is not transmitted over UART. This is because the analog switch IC’s did not work and manual switches were needed to test the basic functionality of the circuit and program. The data is transmitted as a LONG INT and is initialized by the following:UCSR0B|=(1<<TXEN0);UCSR0C|=(0<<UMSEL0)|(0<<USBS0)|(3<<UCSZ00);UBRR0L = 0x0c;UBRR0H = 0x00;Figure 6.5 - Overall program workflow.7 - Mechanical Design7.1 - Research7.1.1 - Research Findings and InterpretationAfter research was conducted into designing a clamp-on ultrasonic sensor and discovering that a clamp-on sensor was not feasible, an integrated design had to be taken. A few different design ideas were discussed but the fundamental design would have to be determined by the transducer selection. One such problem posed was to buy a transducer or make one. The below design was for an exposed piezo-electric element sitting upon a gold washer. The seal was then completed with a cap, and a epoxy seal between the cap and the transducer fitting.Figure 7.1 – Exposed Piezo-electric element design with sealed cap.However after it was determined that it would be best to buy transducer a new method was taken that verified a vacuum seal. It was discovered that standard vacuum seal assemblies existed and could be sponsored through our sponsor company BOC Edwards. These parts were to be purchased and would be modified to meet compliance.7.1.2 – StandardsThe design standard to comply with stated by our sponsor engineer was SEMI-S2. Upon researching through this standard, mechanical piping and sealing would have to be in compliance with ASME Boiler code, section VIII for unfired pressure vessels . However this standard is for pressure vesslesvessels and piping that hold pressure 15 psi or greater. Since this is a vacuum process the gage pressure can never exceed atmosphere pressure 14.7 psi, therefore these standard has specifications that are too great for mechanical design. After discussing this with our sponsor engineer, it was decided that we would not need to design in compliance with either of these standards, and that it would be BOC Edwards responsibility to use these standards when developing this project beyond the scope of this project.However, if the quadrant is to have new flanges these NW flange fittings must be used, since the NW 40 Flanges currently exist on the quadrant. Upon research and finding NW flange assemblies, as many NW rated parts were purchased to enhance this seal.7.1.3 - Overview of the DesignThe design consists of a pair of modified vacuum seal assemblies, complaint with NW parts. Each assembly consists of a welded fitting, based onof a NW 16 long neck flange, which houses a transducer sleeve also having a mating NW 16 flange. A standard Viton o-ring is used to help seal between the welded fitting and the transducer sleeve. To enhance that a seal has been made, a standard NW 16/10 clamp is used to hold the seal together.7.2 - DesignThe ultrasonic transducer is integrally mounted into the quadrant pipe. Mounting through the pipe wall is necessary to measure the gas flow inside due to the large differences in material densities and the resulting energy reflection that would occur if mounted externally. Integral mounting was accomplished by adding ports to the tube wall to accept a sleeve housing the transducer.7.2.1 - Drawings of the QuadrantWe were supplied with sample quadrant tubes that were fully manufactured. We used these tubes in designing the mounting for the transducers and building flow models, which were created using FloXpress within SolidWorks. A detailed drawing of the quadrant section was supplied so that we could obtain accurate dimensions. Refer to the Appendix 1B for a detailed drawing of the original quadrant (provided by Edwards).7.2.2 - Design of the Modified QuadrantTwo machined holes are required in the quadrant tube to mount the transducers and for pass- through of the ultrasonic beams. The holes are placed by setting the appropriate beam reflection angle, and then their spacing is determined by the existing tube geometry. Machining of the holes is done along the axis of beam travel, giving aligned edges with the axis and the proper hole geometry when viewed along the beam axis. The diameter of this hole is nearly the diameter of the transducer sleeve bore so that the two align smoothly when welded.Figure 7.2 - The modified quadrant modifications.7.2.3 - Weld Stub FittingThe fitting designed to permanently attach to the quadrant tube and accept the transducer unit is the weld stub fitting. The prototype model is fabricated with 304-stainless steel. The sealing flange is similar in dimensions to an NW16 weld stub fitting, but the part has a slightly larger outside diameter and bore. The slightly larger size is needed to accommodate an insert to house the transducer that was available to us. Although of larger bore size, it is still able to accept NW16 clamps and O-ring seals but lacks the step required for a centering ring. The face that mates to the quadrant tube is coped to the outside radius of the tube and angled appropriately for aligning the transducer beams. The weld stub fitting seats on the quadrant tube outside face for ease of alignment in manufacturing; alternatively, it could have been inserted into the quadrant tube, but this would pose the problem of setting the depth when welding.Figure 7.3 - All additional parts directly related to the quadrant tube: exploded view.7.2.4 - Transducer Press-Fit SleeveA 304-stainless steel sleeve was designed to house the plastic-bodied transducer and slip fit into the weld stub fitting. The sleeve has a bore designed to press fit the transducer body acting as a sealing mechanism. The outside body diameter is a loose clearance fit with the weld stub fitting this is required for ease of installation/removal due to deformation in the weld stub fitting after being fused to the quadrant tube; a tighter fit may require post-welding machining to retain tolerances. Unfortunately, we were restricted to selecting transducers that were readily available. This made fitting the transducer available into an NW16 format while being able to easily remove the transducer and cap the ports impossible, requiring us to use the same clamping style as NW16 but with oversized tube dimensions. In the future, obtaining a 300-kHz transducer in a smaller housing that incorporates the NW flange would allow us to easily use standard NW16 hardware, only adding the cope or very possibly reducing the size format to NW10.7.2.5 - TransducersThe ultrasonic transducer is an AT300 model from AIRMAR Technologies. Its housing is made from Valox, the brand name for GE’s injection molding resin PET (polyethylene terephthalate), with a white face consisting of Kynar, a brand name for PVDF (polyvinylidene fluoride). A lead of coaxial cable with a BNC connector attached is supplied out the back. The transducer operates at a peak frequency of 300 KHz at a maximum drive voltage of 400 Vp-p. The outside dimensions of the cylinder vary, being narrower near the face and widening near the back end. This dimension variance made it difficult to gauge an adequate press fit. The average diameter was used as the dimension for the press fit.8 - Final Design8.1 - Complete DesignThe final design below is an integrated ultrasonic flow sensor that uses ultrasonic transducers to measure the flow based off the transit time method. The sensor housing comprises NW 16 standard and non-standard parts used to ensure a vacuum seal. The flanged assembly includes an inner sleeve that secures the transducers. Both transducers are hooked up to the electrical box, which contains all of the electrical circuitry.The circuit first generates a pulse that is emitted from the first transducer. A timer is started and then stopped when the pulse reaches the second transducer. This time is stored, and the timer is started again as the second transducer emits a pulse back to the first transducer. As the pulse from the second transducer is received by the first transducer, the timer is stopped and the difference in the calculated. From this time difference the volumetric flow can be calculated. The flow rate is displayed by an LED display on the front of the circuit box and also has the ability to send the flow rate to a computer.8.2 - Sensor Housing AssemblyThe sensor housing is composed of some standard NW 16 and non-standard parts that complete the custom vacuum seal assembly for the flow sensor. When the design needed to be changed from a non-invasive clamp-on sensor to an integrated sensor, the main design concern became creating a vacuum seal. After some research, we found that Edwards produced components that could ensure a vacuum that could be added on to the pipe. This inspired the final design of the sensor housing shown below. (7) Transducer(8) BNC Connector(4) Transducer Sleeve(3) O-ring(2) Weld Fitting(1) Quadrant(5) Vacuum ClampFigure 8.1 - All additional parts directly related to the quadrant tube: exploded view.First, the existing quadrant had to be modified by mdrilling. Because components are mounted at an angle and need a normal projection of that of a circle, two ellipses were machined into the quadrant, with the smallest diameter of 16 mm. A long-necked (70mm) NW16 flange was to be used. However, when the transducers were received, we found that their OD was larger than expected, and a custom weld fitting was made out of 304-stainless steel. The material as it was donated to the design team for use. An inner sleeve that secures the transducer slides into the weld fitment. The transducer is pressed fit into a bored hole in the sleeve. The bottom of the hole is chamfered, allowing lower contact stresses of the inner sleeve where it makes contact with the transducer as well as allowing centering of smaller transducers, should they be used in the future.A Viton NW16 O-ring is used between the two flanged surfaces. It was chosen because of its corrosive resistance to the process gasses to be used as well its ability to handle temperatures up to 200 °C. An exterior NW10/16 clamp is used to ensure that vacuum seal is complete. This is also a stock part purchased from Edwards.8.3 - Insulating Transducer Sleeve and Modified Long Neck FlangeWill fill at a later timeA sleeve was machined to accept the transducers we had available to use in the project. The transducer was press fit into the sleeve and this assembly then inserted into the long neck flange.8.4 - Modified Long Neck FlangeWill fill at a later timeA long neck weld stub fitting was specially machined to accept the transducer sleeve. The bore and OD are slightly larger than a standard NW16 fitting. The sealing flange is similar in dimensions to NW10/16 and will accept clamps and seals for those fittings but lacks the groove for a centering ring. The final part was machined from 304 stainless steel round stock which was donated to the project by Timberland Tool and Die.8.54 - Electrical HousingAll circuitry is housed within a single container. Originally Aan off-the-shelf plastic injection-molded two- piece project box was purchased and further modified to house the custom circuit components. A container of reasonable dimensions was selected, and this posed a soft constraint on sizing of circuit boards and all elements mounted through the container.Solid modeling of the box was combined with models of all components that would require mounting cutouts. Placement of certain components such as the BNC connectors was dependent on their placement on the circuit board whereas other components were allowed to “float,” such as the LCD display array, RS232, and power switch. For prototyping purposes, cutouts were designed for ease of manufacturing while retaining cleanliness, closure, and an ergonomic result.The final circuitry housing was laser cut and formed from 18ga stainless steel. A solid model of the housing was created and the cutouts for the fan, power switch, LCD display and all other ports were all precisely positioned. Laser cutting and CNC forming produced a housing that fit the existing circuit boards and components well. Figure 8.1 - Unmodified circuit boxManufacturing drawing for : bottom section of custom circuitry housing..Figure 8.2 - Unmodified circuit box: top sectionDrawing of housing cover plate.8.5 - Main Circuit Schematic8.6 - 7-Segment Circuit Schematic9 - Testing and Evaluation9.1 - ElectricalThe testing of the electrical circuit was slow going. At each step we encountered problems that were quite difficult to fix since the PCB was small. It is also difficult to modify an existing PCB that does not have any places to attach wires or jumpers.Figure 9.1 - This was our first test of the 7-segment program.Figure 9.2 - Complete setup; a power supply was designed specifically for this project.Figure 9.3 - This is the FIRE pulse from the TDC-GP2 chip; 10 pulses as expected.9.2 - Proving the Accuracy9.2.1 - Rotameter Comparative TestingA Rotameter style airflow meter was used to compare flow results obtained from the ultrasonic sensor. The Rotameter is a variable area flow sensor, using a weight that is lifted by flow in a tube of increasing diameter. As the weight rises, more air can flow past, and the weight maintains its elevation at constant flow rates. Only flow rates within the project outline were tested (10–100 SLM).9.2.2 - Differential Pressure ReadingsIn addition to measuring flow rate, pressure loss data was recorded usingover the designproject flow rates. This data may be useful in calibrating the ultrasonic sensor and in predicting flow properties. The image belowbelow image shows the test setup which consists of PVC piping and connections, both a quadrant with the integrated housing, and plain unmodified quadrant, Rotameter flow meter, and differential pressure sensor. Differential Pressure SensorPressure TapsFlow MeterFigure 9.1 – Image of test setup to measure pressure drop.The test procedure was to use a regulated air flow and measure that the air flows within our design criteria. To do this, a quadrant is selected and subjected to flow rates ranging from 0-3.4 CFM (0-100 SLM). For each measured flow rate reading, a, a differential value is also measured and tabulated. This process wasis repeated several times on the quadrant with and without the sensor. This information was used to see the effects of flow rate versus pressure drop to see if the sensors implementation caused unwanted pressure drop.be9.2.3 - ResultsThe results from the pressure tests show that very little pressure drop occurs due to the sensor’s presence. The below result shows pressure drop measurement between the un-modified quadrant and the quadrant with a built in sensor. The below plot compares all of the data points taken from trials 4-10, testing with and without sensor modification. Trials 1-3 were disregarded due to leaks in the testing apparatus.The increase in pressure drop was quite low as expected, and thus sensor’s impact cause minimal pressure drop in the pipe with the flow. From this data, the design criterion of minimal pressure drop due to the sensor’s implementation was achieved.10 - Future ConsiderationsFuture development of this project is highly recommended. The bulk of the work has been completed. Further modification would require that a detailed plan be provided at the beginning of the capstone project because of the lead times of some of the required components. Furthermore, the sensors should easily be modifiable for application in a working environment. Testing is likely to take a significant amount of time due to the complex synchronization of four sensors into one board. Also, mechanical stresses due to the high heat and corrosive environment will require extensive testing to ensure that the maintenance and life intervals will be met.10.1 - ElectricalFuture designs will use a four-layer PCB to allow for better path routing and thicker paths for certain features, such as the power supply to the chips and the communication to/from the 7-segment display PCB. Furthermore, the goal for Edwards is to have four sensors, with each monitoring the flow through a quadrant.There are many ways to accomplish this. If the high sampling rate is encouraged, then a simple replication of the current circuitry will be done. Essentially, the microcontroller will control four sensors concurrently and wait for the time difference to be sent from each of the four TDC-GP2 chips. Significant modification to the programming would need to be done to incorporate the extra sensors. However, if the sampling rate can be reduced to only 250 samples per second, then the only modification needed would be to replicate the amplifier/receiving circuitry.A significant upgrade in the display would be necessary so that a user could easily see the flow rate. This would best be done by a larger LCD display that could display several lines of information on the same screen. Possible touch sensitivity could be implemented to control the calibration of each circuit independently. Also, a user could manually adjust the sampling, RS232 data output rate, and other functions within the circuit based on the customer’s needs.10.2 - MechanicalRedesigning of the transducer mount to reduce the affect on gas flow would potentially benefit the sensor operation and component lifetimes. Although completely external mounting of the transducers may not be possible due to acoustic impedance mismatch, it may be possible to create a transducer that would sit flush with the inside of the pipe surface when installed.Another area of improvement is the interface between the transducer and gas. Currently, the fixture angle and flat face of the transducer combine to create a cavity into which the gas can deposit material and degrade sensor performance. Redesigning the mounting fixture to be the dimensions of an NW10 weld stub would help reduce this cavity but not eliminate it. Further redesign of the transducer itself would be required.10.3 - TransducersThe transducer used in the prototype design is a readily available model that was donated to us. Custom design of the transducer housing with a specific operating frequency would require a long lead time and production of many units to mitigate costs per unit. If time was available, many improvements to the project could be made.First, incorporating the NW flange into the PET injection-molded body of the transducer would reduce parts, cut production cost, simplify replacement, and allow for smaller package size. This could enable the use of NW10 dimensions and further minimize space required around the quadrant tube.The second improvement is to the transducer face. A curved transducer face could be designed so that, when installed, it is flush with the inside pipe wall, eliminating the cavity and not impeding flow. This would require extensive research and may focus the ultrasonic beam undesirably, but it may be possible using new materials such as PVDF (polyvinylidene fluoride). This is a flexible piezoelectric material that AIRMAR currently uses in transducers.11 - ConclusionWe feel that the complexity and time required for this project were underestimated. In particular, the electrical aspects of the project were found to be very complex and not feasible within the time frame allowed for the senior capstone project.As a whole, we all enjoyed the project and the fact that it was composed of Mechanical and Electrical Engineering students. We learned valuable tools and techniques to interact with other disciplines of engineering and convey information to our sponsor.12 - References[1] Edwards Training PowerPoint, Edwards BOC, 2010[2] Serria Instruments, Product overview; 2010 [3] ‘The Coriolis Measuring Principle’, Endrauss and Hauser, 2009, - AppendixAppendix A - Bill of MaterialsBill of Materials for Mechanical ComponentsMaterialQTYUnit PriceTotal CostStockSupplierPart #316L SS Tubing 16mm1ftDonatedMedalion MetalsNW Flange Blanks2$3.67 $7.34 YEdwardsC10512366O-Ring Viton(5 pack)2$3.82 $7.64 3-4 wksEdwardsH02124013NW 16 70mm Long Flange2$10.33 $20.66 3-4 wksEdwardsC10512316NW 10/16 Clamping Ring SS2$1.11 $2.22 YEdwardsC10512401Additional MaterialO-Ring NW10 w/ Centering Ring2$1.70 $3.40 YEdwardsC10511395O-Ring NW16 w/ Centering Ring2$1.88 $3.76 YEdwardsC10512395NW 16 Flange Blank2$3.81 $7.62 3-4 wksEdwardsC10511366NW 10 70mm Long Flange2$17.15 $34.30 3-4 wksEdwardsC10511316304 Stainless Roundstock3ftDonatedTimberland Tool & DieOutsourced WorkFab - Welding0.5 hrsDonatedHolland FabricationFab - Circuitry Housing1$31.00 GK MachineFab - Machining8 hrs~$400J&J PrecisionMaterial QTY Unit Price Total Cost StockSupplier Part No.316L Tubing – OD = 16mm 1 ft Medalion Metals NW Blank Flange SS 2 $ 3.67 $ 7.34 YEdwards C10512366O-Ring Viton (5 pack) 2 $ 3.82 $ 7.64 3-4 wksEdwards H02124013NW 16 70mm Long Flange 2 $ 10.33 $ 20.66 3-4 wksEdwards C10512316NW 10/16 Clamping Ring Stainless 2 $ 1.11 $ 2.22 YEdwards C10512401???Added itemsCentering Ring with Oring NW102 $ 1.70 $ 3.40 Y?C10511395Centering Ring with Oring NW162 $ 1.88 $ 3.76 Y?C10512395NW Blank Flange SS NW162 $ 3.81 $ 7.62 3-4 wks?C10511366NW 10 70mm Long Flange 2 $ 17.15 $ 34.30 3-4 wks?C10511316Bill of Materials for Electrical ComponentsPart NamePart NumberLocationQuantityPrice (each)Totals:ATmega325P-AU556-ATMEGA325P-2$6.99$13.981:40 1MHz DC Transformer673-4$9.35$37.40High Speed OpAmp595-5$1.87$9.35Analog Switch SPDT700-5$2.42$12.107 Segment Display604-SA03-11PBWA/12$2.40$28.80300V-1A Diode863-24$0.18$4.321n4148 Diode512-10$0.03$0.30Power NMOS512-5$1.72$8.60100nF Ceramic Cap581-50$0.08$4.0022pF Ceramic Cap80-8$0.09$0.7215pF Ceramic Cap80-8$0.07$0.5610uF Tantalum Cap80-8$2.10$16.80.0047uF Ceramic Cap80-4$0.32$1.28100pF Ceramic Cap80-4$0.09$0.364700uF Electrolytic Cap647-2$1.39$2.7815k ohm Pot652-3386P-1-2$2.06$4.121M ohm Pot652-3386F-1-2$1.28$2.56100 ohm 1W Resistor594-5$0.16$0.801k ohm 1W Resistor594-8$0.16$1.281k ohm Resistor260-1.0K-20$0.04$0.8010k ohm Resistor260-10K-20$0.04$0.80150 ohm Resistor292-150-6$0.04$0.24470 ohm Reisistor652-CR0805FX-6$0.05$0.3036 ohm Resistor260-36-20$0.04$0.80Female BNC571-5227161-4$1.80$7.20Male BNC523-31-320-4$1.44$5.76Power Switch611-1$2.30$2.303mm Blue LED's604-WP7104QBC/5$0.29$1.45BCD to 7-seg Decoder595-3$1.48$4.44100 ohm Resistor292-100-10$0.04$0.40220 ohm Resistor260-220-10$0.04$0.400 ohm Resistor (jumpers)652-CR0805-J/-50$0.05$2.505.1 ohm Resistor (5 ohm)260-5.1-40$0.04$1.607-seg connector female538-51353-5$0.38$1.907-seg connector male538-87831-5$0.99$4.95Comparator700-2$2.09$4.18Extra BNC's523-4$1.59$6.36regulator595-2$0.60$1.20opamp579-MCP6023-E/5$1.16$5.80sockets538-50394-60$0.12$7.20header538-87568-4$2.90$11.60max232595-2$0.94$1.88rs232601-40-2$0.54$1.087-seg BJT's610-12$0.81$9.72AC adapter - 24VDC553-WDU24-1$15.14$15.14DC Power Jack502-1$1.53$1.5324V-12V Regulator511-1$0.56$0.5612V-5V Regulator511-1$0.62$0.6212V-3.3V Regulator511-1$0.53$0.53Programmer556-1$34.00$34.00Crystal Oscillator - 16MHz815-ABL-16-2$0.39$0.78Power Switch611-1$2.30$2.30RS232 Interface IC595-1$0.94$0.94TDC-GP2John Menteith (ACAM Product Manager) 513-583-94912$32.85$65.70To order the TDC-GP2, call John Monteith (see the attached quote)??$0.00Serial Input Connectors571-3$2.10$6.30JTAG pin header798-A3C-10P-3$1.48$4.44????$367.81????Appendix B - Manufacturing InstructionsAppendix 1B - Original Quadrant Drawing from Edward’sAppendix 2B - Modifications to the QuadrantAppendix 3B - Transducer FittingsAppendix 4B - Electrical Housing BoxAppendix C - The Program/*-----------------Capstone Project Programming Code-----------------------*/#include <avr/io.h>#include <avr/pgmspace.h>#include <avr/interrupt.h>#include <util/delay.h>#define NOP delayfunction();#define spidisable PORTB|=(1<<PORTB0);#define spienable PORTB&=(0<<PORTB0);void tdcstate(void);void spisend(unsigned long int, int);void timemeasure(void);void timeaverage(float,float);void flowrate(float);void SWITCH(int);void leddisplay(int long);void DATACONVERT(float);void time0init(void);void uart_init(void);void uarttransmit(long int);void GP2_init(void);void communicationtest(void);void hardwaredelay(void);volatile unsigned long int datat,count1; volatile unsigned long int datar; volatile float timeup,timedown,delay;int count=0;unsigned long int FLOWRATE;/*for testing*/ float timeavg=0, timesum=0; volatile int a=0; int led_buf[5];unsigned long int comtest;volatile long int temp[5]; long int ledtestnum;void delayfunction(void){ for(int i=0;i<2;i++) { asm("nop"); }}/*spi initialization read and send data*/ //dont forget clear datar after use it!!!!void spiinit(void){ DDRB|=(1<<PB1)|(1<<PB2)|(1<<PB0)|(0<<PB3); SPCR|=(1<<SPE)|(1<<MSTR)|(0<<SPR1)|(1<<SPR0)|(1<<CPHA)|(0<<CPOL)|(0<<DORD); spidisable;}/*---------------TIME0initialization----------*/ void time0init(void){ DDRG|=0x1f; DDRA|=0x0f; TCCR0A=0b00111101; //CTC,1024,Set OC0A OCR0A=0x0f; //top value TIMSK0=0b110; //OCIE0A}/*-----------UART_initialization-------*/void uart_init(void) { UCSR0B|=(1<<TXEN0); //UCSR0A|=(1<<U2X0);//?¨????±??? UCSR0C|=(0<<UMSEL0)|(0<<USBS0)|(3<<UCSZ00); UBRR0L = 0x0c; UBRR0H = 0x00; }//volatile!!?????±??±???,????!"for, switch, etc."void spisend(volatile unsigned long int data,int bitlength) { spienable; datat=data;if(bitlength==8){SPDR=datat;while(!(SPSR&(1<<SPIF)));//temp[i]=SPDR;datat=data;}if(bitlength==32) { for(int i=0;i<4;i++) { datat<<=i*8; SPDR=datat; while(!(SPSR&(1<<SPIF))); temp[i]=SPDR; //temp[i]=0xabcd; //test datat=data; NOP; NOP; } }if(bitlength==40) { for(int i=0;i<5;i++) { datat<<=i*8; SPDR=datat; while(!(SPSR&(1<<SPIF))); temp[i]=SPDR; datat=data; NOP; NOP; } } spidisable;}void spiread(int readbitlength) // "sendtype" 5: COM_test 4: state 1:result{ if(readbitlength==8) { datar=temp[1]; } if(readbitlength==16) //State adjustment 16bits { datar=(temp[1]<<8)+temp[2]; } if(readbitlength==32) { datar=((temp[1]<<24)+(temp[2]<<16)+(temp[3]<<8)+temp[4]); }}/*---------Time_Measurement_Function----------*/void timemeasure(void){ if(count==0) { //SWITCH(0); spisend(0x01,8); }else { //SWITCH(1); spisend(0x01,8); } while((PIND&(1<<PD1))); //should be there tdcstate();}/*-----------TDC_STATE_JUDGEMENT_FUNCTION--------*/void tdcstate(void){ spisend(0xB4,32); spiread(16); if((datar&0x200)==0) //time data overflow 10's bit set { spisend(0xb0,32); //read_reg0 OPCODE; //maybe b1,b2spiread(32); //read 32bits time data;if(count%2==0) {timeup=(float)datar;}else {timedown=(float)datar;} count++; datar=0x0; if(count==2){count=0;} }else { datar=0x0; //else don't store the data } }/*---------Time_Average_Function----------*/void timeaverage(float TIMEUP,float TIMEDOWN){ if(count%2==0) { timesum+=(TIMEUP-TIMEDOWN); timeavg=(timesum)/(count1); //cal timeavg if(count1==1000) { count1=0x0; flowrate(timeavg); } count1++; } }/*---------flowrate_calculation_Function----------*/void flowrate(float TIMEAVG){ float flowrate; flowrate=TIMEAVG; FLOWRATE=(long int)flowrate; //for testing}/*---------Switch_Function----------*/void SWITCH(int direction){ if(direction==0) // updirection direction=0 { PORTC=(1<<PORTC0)|(0<<PORTC1)|(0<<PORTC2)|(1<<PORTC3); } if(direction==1) { PORTC=(/*0*/0<<PORTC0)|(1<<PORTC1)|(1<<PORTC2)|(0<<PORTC3); }}/*----------convert float to long int----*/void DATACONVERT(float data){ long int datadisplay; datadisplay=(data*100)+0.5; leddisplay(datadisplay);}/*------------store data to buffer for display------------*/void leddisplay(long int data){ long int NUM1; led_buf[4] = (data / 10000);/*??????????????????????????*/NUM1 = data % 10000;led_buf[3] = (NUM1 / 1000);NUM1 = NUM1 % 1000;led_buf[2] = (NUM1 / 100);NUM1=NUM1%100;led_buf[1] = (NUM1 / 10);led_buf[0] = (NUM1 % 10);}/*-----------Time0interrupt_for_dynamic_scane_display---------*/SIGNAL(TIMER0_COMP_vect){ if (a == 4) {/*??????*/PORTG=(1 << 0); PORTA=led_buf[4];}if (a == 3) {PORTG=(1 << 1);PORTA=led_buf[3];}if (a == 2) {PORTG=(1 << 2);PORTA=led_buf[2];}if (a == 1) {PORTG=(1 << 3);PORTA=led_buf[1];} if (a == 0) {PORTG=(1 << 4);PORTA=led_buf[0];} a++; if (a > 4) {/*???????í*/a = 0;} // _delay_ms(100);}/*----------uart_data_transmit--------*/void uarttransmit(long int data){ for(int i=0;i<4;i++) { UDR0=data>>(24-i*8); while(!(UCSR0A&(1<<UDRE0))); }}/*-----tdc-gp2 initialization---------*/void GP2_init(void){ long int REG0,REG1,REG2,REG3,REG4,REG5; long int PU=0X50,Init=0x70; REG0=0X80AB5668; //80AB5668; REG1=0X81214200; REG2=0X82E03200; REG3=0X83200000; REG4=0X84203400; REG5=0X82000000; spisend(PU,8);//???????? NOP; spisend(REG0,32); NOP; spisend(REG1,32); NOP; spisend(REG2,32); NOP;// spisend(REG3,32); NOP; //spisend(0xB0,40);// spisend(REG4,32); NOP;// spisend(REG5,32); NOP; spisend(Init,8);}/*---------TDC-GP2_COMMUNICATIONTEST--------*/void communicationtest(void){ spisend(0xB5,32);NOP;spiread(8);comtest=datar;datar=0x0;}/*-------------------hardware_delay---------------*/void hardwaredelay(void){for(int i=0;i<2;i++) { timemeasure(); } delay=(timeup-timedown);}int main(void){ DDRC=0xff; time0init(); sei(); uart_init(); spiinit(); GP2_init(); //SWITCH(0); communicationtest(); hardwaredelay(); while(1) { timemeasure(); timeaverage(timeup,timedown); uarttransmit(FLOWRATE); DATACONVERT(FLOWRATE); } return 0;}Appendix D - CalculationsTime of FlightThe time of flight is calculated by using the Pythagorean Theorem of right triangles to calculate the distance that the gas will travel. Further, we were also able to calculate the Doppler shift caused by the gas flowing through the pipe. Figure 5.4 will be used to demonstrate the calculations.L= d1sinθ→Ltotal=2LWhere d1 is the inside diameter of the pipe, and L is the length from the inside of the pipe to the center point between the two transducers. Thus, the time of flight is then given by the distance of travel (Ltotal) divided by the speed of sound in air (340m/s).In order to calculate the time of flight of the ultrasonic pulse, you must consider the horizontal velocity vector that the gas adds or subtracts to the sounds horizontal velocity. The red in the figure below shows the final triangle of the ultrasonic pulse, and the black triangle is the original velocity vectors of the ultrasonic pulse.Figure D1 - Shows the velocity vectors of the ultrasound and the gasWhere Vsy and Vsx are the vertical and horizontal, respectively, vector components of the ultrasounds velocity. Vso is the straight path velocity of the ultrasound, Vg is the velocity of the gas, and Vds is the final velocity of the ultrasound traveling through the pipe.Vsy=Vsound,airsinθ, Vsx=Vsound,aircosθVds is then expressed as: Vds=Vsy2+Vsx+Vg2. To calculate Vus for an upstream pulse, you would subtract Vg from the equation.Time of Flight=Ltotal Vsound,air=LtotalVds or LtotalVus→?t=tup-tdownFurthermore, it is appropriate to calculate the Doppler shift the ultrasonic pulse experiences during its transit. This is done by first calculating the time of flight of an up or down stream pulse. Typically, the Doppler shift is negligible and will not affect the performance of the sensor.Doppler Shift= ?x=Vg*LtotalVdsVolumetric Flow RateOnce the difference in the time of flights is calculated, an equation can be used which is independent on the geometry of the pipe.Volume= Ltotal2cosθ*?ttup*tdownReflection and Transmission CoefficientsBelow is the Matlab code which was written to calculate the reflection and transmission coefficients which showed that a clamp on sensor was not feasible when gas is flowing through a stainless steel pipe.clear allclcformat compactformat short%format loose%Incidence Angle: off the pipe wall - in degreestheta_i = 30;%Temperature of Gas: in CelciusTg = 165; %Converting Gas Flow rate from slm to cm/sVg_ini = 70; %in slmPg = 1; %pressure of gas in atmDg = 1.429; %density of gas in PaVg_calc = 355.2; %converted velocity in cm/s%Calculate Velocity of Ultrasound in the Gask = 1.3806504*10^-23; %Boltzman's Constant J/KTg_abs = Tg + 273.15; %Temperature in Kelviny = 7/5; %Adiabatic Index (for all materials)mF = 18.998404; %mass of Fluorine atom in amumCl = 35.4527; %mass of Chlorine atom in amumO = 15.9994; %mass of Oxygen atom in amu%Find mass of gas mixture% 60% Fluorine, 30% Chlorine, 10% Oxygenmg_kg = (.6*mF+.3*mCl+.1*mO)*1.66054*10^-27;Vus_gas = sqrt((y*k*Tg_abs)/(mg_kg));Bss = 2.5*10^9; %Bulk Modulus of 316l in Pa (N/m^2)Dss = 1430; %Density of 316l in kg/m^3 (remains constant for all T)Vus_ss = sqrt(Bss/Dss); %Calculate the Transmission Angle in degreestheta_i_norm = (90-theta_i)*pi()/180;theta_t = asin(Vus_gas/Vus_ss*sin(theta_i_norm));Transmission_Angle_rad = theta_t;%Characteristic Acoustic Impedance of the Mediums:Zgas = (.6*1.696+.3*3.214+.1*1.429)*Vus_gas; %in ohmsZss = Dss*Vus_ss; %in ohms%Transmitted Powernum = 4*Zgas/Zss*cos(theta_t)/cos(theta_i_norm);den = (Zgas/Zss+cos(theta_t)/cos(theta_i_norm))^2;TP = num/den;RP = 1 - TP; %Values to Display:Velocity_of_Ultrasound_GAS = Vus_gasVelocity_of_Ultrasound_316SS = Vus_ss%Off of the pipe wallUltrasound_Incidence_Angle = theta_iUltrasound_Reflection_Angle = theta_i%Off of perpendicular to the pipe wallUltrasound_Transmission_Angle = theta_t*180/pi()Characteristic_Acoustic_Impedance_GAS = ZgasCharacteristic_Acoustic_Impedance_316SS = ZssTransmitted_Power_Percentage = 100*TPReflected_Power_Percentage = 100*(1-TP)Flow AnalysisPurpose: To examine the flowrates and determine fundamental values such as Reynolds number so that pressure loss calculations can be estimated.Solution:Specified flow properties and valuesMeasured ConditionsTsTaPsPa293308760766Literft^310.035314667Pipe Diameter (m)0.028Pipe Area (m^2)0.000615752Actual fFlowrate from sStandard Fflowrate (Figliola & Beasley pg419)Qa=Qs×TaPsTsPaFlow Velocity; Munson, Young, OkiishiV=QaApActual Flowrates (SLM, cfm)???VelocitiesQsQaQaQaVSLML/mm^3/s(cfm)(m/s)1010.431.74E-040.370.282020.863.48E-040.740.563031.295.21E-041.100.854041.726.95E-041.471.135052.158.69E-041.841.416062.581.04E-032.211.697073.011.22E-032.581.988083.441.39E-032.952.269093.871.56E-033.312.54100104.301.74E-033.682.82Reynolds NumberRed=4×QaπdνReynolds Number Results:Reynolds #Kinematic Viscosity (ν) Air @ 50°C (m^2/s)0.0000179Min ReMax Re441.58643584415.864358Conclusion:The results for the Reynolds numbers show that the flow is laminar at its slowest rate (Re=442). At the highest flowrate (Re=4416) the flow is in what may be a turbulent transitional zone (Re=2000 to 6000). Likely due to the piping previous to the quadrant tube the flow will become turbulent at the low end of the transitional range.Further equations are included that describe the pressure loss.Laminar Friction Coefficientλ=64ReTurbulent Friction Coefficient1λ=-2log2.51Re×λ+KD0.269Pressure Drop from Piping?P=λLDρ2?2Pressure Drop due to Obstruction?P=ζρ2?2Heat Transfer AnalysisPurpose: To see if PEEK plastic could be used as an insulating transducer sleeve for a future design.Solution:First determine the amount of heat transfer occurring in a hollow welded fitting. The point in which heat transfer will looked at is where a transducer would be placed with respect to the welded fitting:Q=T1,tube-T2,gasRTotWhere ,RTot= R1+R2 ; R1=ln?(r2r1)2πkL ; R2=1hgas2πr1L Letting the following variables equal the following;T1 = 180 °C ; T2 = 80° C; r1 = 14 mm; r2 = 15.5 mm; L = 40 mm, k = 13.4 W/m2 – Kh = 10 W/m2 – KCalculating the following variables results in the following;RTot= ln?(15.7514)2π13.4(0.04) + 1102πln(014/.012)(0.04) = 2.61Q=180-802.61 = 38.224 WNow the heat loss will be found by placing a 1mm walled tube inside the welded fitting with a 1 mm air gap.A new resistance will now be introduced called R3.R3=ln?(1210)2π0.23(.04) = 3.15Therefore new Total Resistance = 2.61 +3.15 = 5.76Q=180-305.76 = 26.042 WConclusion:This approach shows that insulating sleeve made of PEEK plastic would be able to take the temperature down to a design temperature of 30° C from 180°C occurring in the welded fitting. However this approach used a couple of assumptions being the temperature of the gas at a point inside of the welded fitting as well as the coefficient of convection, h. Although these are reasonable assumptions, it would be best to experimentally determine both of these values. However letting these values roam, as long as the heat loss is not less than 26.042 in the analysis of the hollow weld fitting the maximum temperature of the ambient gas can be is 112 °C.Temperature Distribution of Transducer Sleeve (AxisymmetricAxisymetric model)Maximum temperature = 180 °C (450 K)Minimum Temperature = 40 °C (310 K)Mechanical AnalysisThermal Strain of transducer sleeve and welded fittingEq (3-61) (Shigley) ?=α?T = ((17x10-6)/C°)(150 C°) = 0.0026α = Coefficiant of Thermal Expansion of Stainless steel (1/C°)? = Strain (mm/mm)?l=l0 ? = (0.0026)(1.54mm) = 0.0031mm?l = Change in length (radius) (mm)l0 = Original length (mm)Tangential StressWelded fitting – (σt)max= p(davg)2t = 14.6 psi(0.0195m)2(0.00153m)* 6894 Papsi= 640 KpaTransducer Sleeve - (σt)max= p(davg)2t = 14.6 psi(0.0175m)2(0.00153m)* 6894 Papsi=570 Kpap = Applied pressure (psi)davg = average diameter (m)t = wall thickness (m)Press Fit (Sleeve and Transducer)(Eq 3-56) (Shigley) – p=δR[1E0r02+R2r02-R2+υ0+1Eiri2+R2ri2-R2+υi δ = radial interference (m), R= nominal radius (m), ‘o' and ‘i' are subscripts that refer to inner and outer components, E = Elastic Modulus (GPa), r = radius of component (m) and υ = poisson’s ratioLetting,δ = 0.00008m, R = 0.006m, ro=0.00585m, ri = 0.00615m, Eo = 190 GPa, Ei = 2 Gpa, υ0 = .3, υi= .37P = 440 KPaFEA ModelsTransducer Sleeve (Axisymetric Model)–Boundary Conditions and LoadsLoads – Pressure = 14.6 psi (1 x 105 Pa)Boundary Conditions – Contact surface of the transducerDisplacement modelVon Mises Stress modelConclusion:Max Displacment = 0.00004 mmMax Stress = 900 KPaAlthough the models visually look quite dramatic, the numbers state that displacement and stress are none factors to failure as expected. These models can be useful after field testing to observe certain areas for pre-mature failure. These models indicate that a buckling of the wall is the most likely place for failure to occur. Transducer (Axisymetric model) – Boundary Conditions and LoadsStress ModelDisplacement ModelConclusionMax displacement = 0.0004mmMax Stress = 1.98 MpaLarger stresses occur at the middle axis. This is due to the seating surface and material the transducer is made out of. This point is also well below a failure point (80 GPa). The displacment is also negliable. Appendix E - Top Level ResearchThermal Sensing TechnologySummary:After researching the following products it appears that one of the following sensors may work under the design constraints which is the “Fox Thermal Instruments FT2”. There claim is pressure is negliable, and an email has been sent asking for a more detailed explanation. The last product from Serria Instruments claims low pressure drop which can be negliable depending on a flow rate.Two main designs were found. The first was sticking an insert able probe into the flow of the pipe. This technology is used with the FT2 and there claim was that the pressure drop was negliable. As mentioned above this is being investigated further. The second design was uses a capillary tube to measure a small portion of the flow. Serria Instruments also claims a small pressure drop due to this design. For a video of the technology go to this page, main problem with all these flow sensors found is that the either do not meet the temperature range criteria or marginally meet it. It is not clear if this is due to the material selection, or the sensing instrument. Another problem is that this means that the pipe will need to cut or manipulated to use any of these flow sensing technolgy.Thermal Flow Sensors:Flow Sensor #1 – Fox Thermal Model FT2 Gas Mass Flowmeter and Temperature TransmitterPRODUCT DESCRIPTION The Fox Model FT2 Thermal Gas Mass Flowmeter and Temperature Transmitter measures two important process variables in one rugged instrument. The FT2 measures gas flow rate in standard units without the need for temperature or pressure compensation. It provides isolated 4 to 20 mA and pulse outputs for flow rate and a 4 to 20 mA output for process gas temperature. You choose the flow rate and temperature engineering units. An optional on-board 2 x 16 characters, backlit display is available to view flow rate, total, elapse time, process gas temperature, and alarms. The display is also used in conjunction with the Configuration Panel to configure flowmeter settings such as 4 mA and 20 mA for flow rate and temperature, pulse output frequency scaling, pipe area, zero flow cutoff, filtering (dampening), display configurations, diagnostics and high or low alarm limits. If you prefer, you can view measurements and set parameters with an optional Palm PDA instead of the on-board Display and Configuration Panel. The FT2 is available in insertion and in-line models. The insertion meter is easily installed by drilling a ?"hole in the pipe and welding on a ?" NPT coupling. A Fox supplied compressing fitting secures the probe in place. The in-line model is available in ?-inch to 6-inch sizes and include built in flow conditioners that eliminate the need for long, straight pipe runs. The meter can be ordered with flange or NPT end connections. Both models are supplied with 316 stainless steel wetted materials standard or Hastelloy C-276 as an option. RS232 for connecting a Palm PDA or computer; RS422/RS485-Modbus, Profibus-DP, DeviceNet and Ethernet give the FT2 flexible communications capability. The FT2 is an advanced Thermal Mass Flowmeter and Temperature Transmitter for your most challenging gas flow measurement applications. Common Gases: Air, ammonia, argon, biogas, butane, chlorine, compressed air, carbon monoxide, carbon dioxide, digester gas, ethane, ethylene, helium, hydrogen, methane, mixed gases, natural gas, nitrogen, oxygen, propane, and many more. (1)Measures gas flow rate in SCFM, NM3H, Kg/Hr, & many more Measures process gas temperatureOutputs: 2 x 4 to 20 mA - one for flow rate and one for temperature; pulse output for flow/total RS232 for connecting a Palm Handheld or computer; RS422/ RS485-Modbus, Profibus-DP, DeviceNet & Ethernet Insertion and in-line models All welded, 316 SS sensor construction; Hastelloy Low-end sensitivity - leak detection Microprocessor based, field programmable On-board 2 x 16 character, backlit display with configuration panel to view/set readings and parameters Palm handheld terminal available to view/set readings and parameters when on-board display & configuration panel is not ordered NEMA 4X enclosure; designed for Class I, Division 2, Groups B, C, & D hazardous areas NIST traceable calibration; CE approvedNegligible pressure drop No moving partsOther Specifciactions of Note:Accuracy within + 1% for flowResponse time 0.9 secondsCons: Some Pressure Drop – Claimed to be negliable Welding and Pipe manipulation requiredPros: Is available in Stainless Steel and Hastelloy Accuracy within acceptable range Measure both Flow and TemperatureFlow Sensor #2 - EL-FLOW?? Laboratory style, digitalGAS Mass Flow Meters / Mass Flow ControllersGeneral ()EL-FLOW? Series Mass Flow Meters/Controllers are thermal mass flow meters of modular construction with a 'laboratory style' pc-board housing. Control valves can either be integrally or separately mounted, to measure and control gas flows from:? lowest range 0,02...1 mln/min? up to highest range 25...1250 ln/minThe control valve design of the mass flow controller distinguishes itself from competitive designs in its truly modular construction and it can be field replaced or changed by the user without any adjustment. The standard valve is normally closed and is available up to Kv-values of 1.5. Normally opened valves can also be supplied. Patented constructions enable us to handle high flows and/or pressures at differential pressures up to 400 bar in the EL-FLOW? program, which is unique."Multibus" Digital Mass Flow Meters/ControllersIn some applications there are more requirements for a modern instrument than analog-based technology can offer. Examples of these requirements are self-diagnostics, alarm and counter functions, digital communication and remotely adjustable control settings. These requirements can only be met with a digital based instrument. Therefore Bronkhorst High-Tech developed a fieldbus based pc-board for their mass flow metering and control solutions. The latest digital instruments offer great flexibility thanks to the "multibus" concept, thanks to which the instruments can be equipped with on-board interface board with DeviceNetTM, Profibus-DP?, Modbus?or FLOW-BUS protocol. More information about digital "multibus" communication.FeaturesHigh accuracy (standard?0,5% of Rd plus 0,1% of FS)? Fast response (down to 200 msec) Pressure ratings up to 400 bar (higher?on request) Electro-chemical polish of all surfaces Compact, modular construction No moving parts Metal sealed and/or down-ported versions available ApplicationsAnalytical/environmental equipment Gas flow monitoring in food, chemical and petrochemical industries Gas consumption measurement in gas distribution systems for internal accounting purposes Detection of gas leakage through objects Semiconductor manufacturing Surface treatment installations Flow capacities (based on air)Mass Flow Meters (MFM);PN100RangesSeries F-110Cmin. 0,02 ... 1 mln/minmax. 0,2 ... 10 mln/minSeries F-111Cmin. 0,2 ... 10 mln/minmax. 0,3 ... 15 ln/minSeries F-111CM (metal sealed)min. 0,1 ... 5 mln/minmax.?2 ... 100 ln/minSeries F-112ACmin. 0,2 ... 10 ln/minmax. 5 ... 250 ln/minSeries F-113ACmin. 2 ... 100 ln/minmax. 25 ... 1250 ln/minMass Flow Meters (MFM);PN200?/ PN400RangesSeries F-120M / F-130Mmin. 0,2 ...?10 mln/minmax. 0,3 ... 15 mln/minSeries F-121M / F-131Mmin. 0,3 ... 15 mln/minmax. 0,4 ...?20 ln/minSeries F-122M / F-132Mmin. 0,2 ... 10 ln/minmax.?6 ... 250 ln/minSeries F-123M / F-133Mmin. 4 ... 200 ln/minmax.?25 ... 1250 ln/minMass Flow Controllers (MFC);PN64 / PN100RangesSeries F-200CV / F-210CV?1)min. 0,02 ... 1 mln/minmax. 0,2 ... 10 mln/minSeries F-201C /?F-211C 1)min. 0,2 ... 10 mln/minmax. 0,3 ... 15 ln/minSeries F-201CM (metal sealed) 1)min. 0,2 ... 10 mln/minmax.?2 ... 100 ln/minSeries F-201AC / F-211AC?1)min. 0,2 ... 10 ln/minmax. 1,4 ... 70 ln/minSeries F-202AC / F-212AC?2)min. 0,2 ... 10 ln/minmax. 5 ... 250 ln/minSeries F-203AC / F-213AC?2)min. 2 ... 100 ln/minmax. 25 ... 1250 ln/minMFCs for high-pressure /high-?P applications; PN400RangesSeries F-230M?3)min. 0,2 ...?10 mln/minmax. 10 ...500 mln/minSeries F-231M?3)min. 6 ... 300 mln/minmax. 0,2 ... 10 ln/minSeries F-232M?3)min. 0,14 ... 7 ln/minmax.?2 ... 100 ln/min1) with small type of valve for normal applications ( Series F-200 / F-201 )2) with pilot-operated valve for high flow applications (Series F-202 / F-203 )3) with vary-p-valve for high (differential) pressure applications up to 400 bar ( Series F-230 / F-231 / F-232 )Note: Mass flow meters ( Series F-110 / F-111 / F-112 / F-113 ) can be close coupled with Control Valves ( Series F-001 / F-002 / F-003 / F-033 ) to constitute a compact mass flow control assembly.Meets the following Design Requirements:Size Requirements: NoMaterial: YesTemperature Range: NoPressure Range: UnconfirmedDoesn’t Obstruct Flow: UnconfirmedAccuracy: YesFlow Sensor #3 – Serria Flow Instruments – Max Track Model 180Max-Trak? Model 180 Industrial Instruments with NEMA 6 / IP67 RatingMax-Trak? is a family of industrial gas mass flow instruments from the company that has been a trusted name in industrial thermal mass flow meters for decades—Sierra Instruments. Max-Trak? will measure and control any clean, dry gas mass flow from 10 to 1000 SLPM, with lower flows and higher flows available upon request.Based upon Sierra’s successful Smart-Trak? line of digital instruments, Max-Trak? adds rugged industrial packaging to popular premium features: ? Dial-A-Gas? multi-gas capability? 316 stainless steel construction? 4 types of analog and 2 types of digital communication ? a wide variety of field adjustable parameters? compact sizes to measure and control flow rates from 10 to 1000 SLPM (higher upon request)When added inside the industrial eclosure,? the internal Compod? programmable control module gives RS-485/MODBUS RTU communications capability to the Model 180.? Compod greatly simplifies basic flow control installations to permit? networking of multiple instruments. Meets the following Design Requirements:Size Requirements: NoMaterial: YesTemperature Range: NoPressure Range: UnconfirmedDoesn’t Obstruct Flow: Maybe – Negliable pressure drop claimedAccuracy: YesThe following video shows how the sensor works; Flow sensor #4 – Thermal instrument company - Model 600-9Summary – ()It's really very simple. Our Thermal Mass Flow Meters operate using a constant temperature system that employs two RTD sensors; one for sensing temperature, and one for sensing flow. The sensor is heated to a precise temperature above that of the fluid passing by. The Fluid conducts heat off the sensor in direct proportion to the mass flow rate. The temperature is used to set the heat on the flow sensor and correct for changes in the fluid temperature. Completely unobstructed flow; Sensor is incorporated into the outer surface protected from the adverse conditions and out of the flow path. Minimal pressure loss; Straight tube design helps maintain steady pressure.Rugged construction; Provides longer life and better performance. Handles a variety of materials; Can be used for liquids, slurries, gases, and homogenous solids. No moving parts to break or wear out. MODEL 600-9 - HOW IT WORKS - No obstruction for better flowSpecifications ?Accuracy: 0.5% of Full Scale or 2% of reading - whichever is better ?Repeatability: Better than 0.2% of reading ?Temperature Extremes: -250°F to 1100°F (-156°C to 593°C) ?Pressure Extremes: Up to 60,000 psi (dependent upon size) ?Response Time: Liquids less than 1/2 second, Gases 1 to 2 seconds typical (4 to 8 seconds worst case) ?Pressure Drop: Negligible - completely unobstructed straight thru flow path ?Flow Element: Construction meets NEMA 4, 7 & 9 explosion proof requirements ?Flow Wetted Parts: 316 S.S. standard, also many exotic metals and coatings available to suit requirements Operating Principle: The Model 600-9/9500 operates using a constant temperature system that employs two RTD sensors: one for sensing temperature, and one for sensing flow. The flow sensor is heated to a precise temperature above that of the fluid passing by. The fluid conducts heat off the sensor in direct proportion to the mass flow rate. The temperature sensor is used to set the heat on the flow sensor and correct changes in the fluid temperature. Thermal Instrument's thermal mass flowmeters consist of one piece; a flow element with a bridge controller, signal conditioner with power supply, flow signal, A/D converter, linearizing network, 4-20 MA DC, flow output, operational flow display, and totalizer.Appendix F - Manufacturing DrawingsAppendix G - Raw DataRaw Data from pressure tests – Trial 1 - With SensorFlow (CFM)Flow (slm)Pressure Drop (in of H2O)1.234.00256.60.01384.90.044113.20.08Trial 2 - With SensorFlow (CFM)Flow (slm)Pressure Drop (in of H20)128.301.234.001.645.30.005256.60.0122.673.60.028384.90.0473.496.30.0734113.20.08Trial 3 - With SensorFlow (CFM)Flow (slm)Pressure Drop (in of H2O)1.439.60.003256.60.0122.467.90.0242.879.30.032384.90.043.496.30.05Trial 4 - With SensorFlow (CFM)Flow (slm)Pressure Drop (in of H2O)1.234.001.851.00.006256.60.0092.467.90.0252.879.30.039384.90.453.290.60.05Trial 5 - Without Sensor Flow (CFM)FLOW (slm)Pressure Drop (in of H2O)1.645.30.0091.851.00.02256.60.0282.262.30.032.467.90.042.673.60.0532.879.30.057384.90.0623.290.60.0773.6101.90.085Trial 6 - Without Sensor Flow (CFM)Flow (slm)Pressure Drop (in of H2O)1.851.00.0152.262.30.0282.879.30.053.290.60.08Trial 7 - Without SensorFlow (CFM)Flow (slm)Pressure Drop (in of H2O)1.851.00.0122.467.90.0332.879.30.0513.290.60.0623.6101.90.084Trial 8 - With SensorFlow (CFM)Flow (slm)Pressure Drop (in of H2O)1.851.00.022.262.30.042.673.60.06384.90.083.496.30.12Trial 9 - With SensorFlow (CFM)Flow (slm)Pressure Drop (in of H2O)1.439.60.0151.645.30.0181.851.00.0252.262.30.0522.673.60.065384.90.0923.290.60.105Trial 10 - With SensorFlow (CFM)Flow (slm)Pressure Drop (in of H2O)1.439.60.0151.851.00.0282.262.30.052.673.60.073.290.60.11 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download