Tuesday, October 29, 2019

Hotplate Wind and Rain Sensor Essay Example for Free

Hotplate Wind and Rain Sensor Essay By keeping the two heated plates at a constant set point temperature it is possible to measure the power necessary to maintain this temperature. The change in power over time needed to maintain this state can be equated with wind removing heat from the plates. We can also evaporate water off one of the plates and measure the work done by the system in evaporation. The further energy required by this plate for water evaporation when measured will constitute a rain sensor. These plates will from now on be referred to as the rain plate or top plate and the wind plate or bottom plate. If the bottom plate is kept dry while being exposed to the wind it can effectively act as a reference plate to evaluate how much extra work is being done to the plate exposed to the rain also. The plates can be seen as a point source in terms of wind exposure since the heated area is small, being ~35mm*65mm. By measuring the work done on the bottom plate from some initial reference point an estimate of the cooling effect of the wind on the plate can be made, e. g. f it takes 10W/hr to maintain a temperature of 25 °C in a dry wind free environment of 20 °C and it takes 2W/hr to maintain a temp of 21 °C in the same environment then in this case if there is an additional 2W/hr used in the system we can say that a wind chill factor of –1 is present. By using a chart of wind chill factors and an equation to calculate the wind chill velocities we can estimate the win d speed. The current standard equation describing wind chill is: Ideally this equation can be solved with a temperature reading derived from the wind plate and compared to a chart of wind chill factors to find a velocity value. The problem with this approach for laboratory purposes is that the equations results becomes unreliable for temperatures above about 10 °C and since all experimental readings were conducted at room temperature i. e. ~23 °C it was difficult to use this method satisfactorily. For the evaporation of rain the latent heat of evaporation or enthalpy of evaporation of water DHvap is tabulated and has been measured as 100 °C40. 657 kJ/mol 80 °C41. 585 kJ/mol Assuming that at 90 ° the value is 41kJ/mol and that the molecular mass of water is 18. 01508 gram/mol then using this value we can say that: If we also take into account the temperature of the Aluminium tray and the heat it passes to the water then we should get an estimate for how long it will take the system to evaporate 1ml of water. For the Al tray the specific heat is taken as 938J/kg/ °K and the temperature drops by about 8 °C when 1ml of water is applied. Using where m = 10. 7*10-3 kg , c = 938 J/kg/ °K , dT = 8 ° giving the energy imparted to the water by the Aluminium trays QAl = 80J By dividing the DHvap of the rainwater by the energy supplied by the plates we should get an estimate of the time it will take to evaporate 1ml of water from the plates. Since the time taken for the plate to reheat to its 90 ° set point value is included in the measurements it is unnecessary to include the heat transferred by the Aluminium plates in the estimates. Thus Since 1g of water  » 1ml of water it’s a good estimate to say this is the time the system will take to evaporate 0. 416mm of rain. The conversion rate of 1ml of water = 0. 416mm rain or 2. 4ml of water = 1mm of rain is based on the dimensions of the Aluminium trays which have a capacity of 12ml and a depth of 5mm. 412s/ml turns out to be a slightly high estimate for the time the system takes to evaporate 1ml of water. This is possibly due to impurities in the Aluminium trays, which can lead to increased thermal emissivity. Also impurities in the water can lead to a lower density and lower latent heat. 3INSTRUMENTATION, HARDWARE AND SOFTWARE 3. 1Hardware The initial hardware in the lab was inadequate although it provided a good copy of what is probably used in similar commercial applications. There were 2 x 2cm2 Aluminium plates which were heated by small power resistors, model RF2222. These were switched by the IRF510 0. 5? rectifier mosfets, which were inset in series with 47K? resistors to reduce overheating. Both circuit loops were wired in parallel to a single power source, i. e. while each switch had its own individual digital logic voltage of 5v or 0v, the hotplates shared one power source. This made it difficult to establish which plates were active and to distinguish them from one another. Another problem with this initial circuit setup was the inadequate size of the hotplates themselves. They were too small to attach the thermocouple devices to and too flat to collect rain in since they were only flat Aluminium. These were replaced with 2 Aluminium trays which had an area of 275 mm2 , an inside depth of 5mm, a total volume of ~12. ml each and masses of 10. 70g for the top tray and 12. 35g for the bottom tray respectively. Figure 1 Detail of the Aluminium Trays and Hot plates Figure 2 shows a block diagram for the final circuit design. Figure 2 Block Diagram for the Hot plate Circuit The first change made was to replace the power resistors with 2 DBK 30W hotplates with 5. 5W/m2 heat output. These are aluminium plates heated with nichrome wire wrapped in a tubular pattern inside them. They are available in a wide range of operating modes, both ac and dc and with various power and temperature ratings. For this project I chose the HP03-1/08-24 model since they have a maximum power drain of 30W and operate at DC voltages between 12—24V. Testing them with a voltmeter showed them to have a variable resistance. They recorded a maximum of 100W at their peak temperatures and 10W at their minimum. The circuits were redesigned so that each loop had its own individual power source. This was done to more accurately measure the activity of each plate. With a separate current and voltage indicator for each plate it was easier to determine which plate was drawing power, how much and when. It also helped overcome problems of inadequate power supply to the plates, which could demand a current in excess of 2A when heating initially due to their low resistance at that stage. The IRF510 mosfet switches in series with a 49KW resistor were later replaced with lower resistance IRFZ34N rectifiers which had a lower resistance of 0. 04? compared to 0. 5W for the IRF510 to reduce heating which was occurring in them in the initial heating phase of the plates when the system was switched on. These were added to the circuit in series with a 3. 9KW resistor. The purpose of these resistances was to avoid overheating in the mosfets causing thermal breakdown. During this phase the power to the plates was a maximum as the plate resistance was lowest ~10?. In addition the power drain was continuous until they reached their set temperature so the amount of time they were switched off was a minimum. Once they reach the set point temperature they then began switching and so the continuous on time for each one was reduced. It was this time difference along with the temperature difference of the plates that was measured in this experiment. The temperature at the Aluminium measurement surfaces was measured by 2 type-j thermocouples, which were connected to the PC via the USB-TC data acquisition unit. Aluminium trays were attached to the hotplates using a nylon clip and screws and the thermocouple devices were attached to the same side of these trays further down to measure the temperature of the Aluminium trays as opposed to the temperature of the hotplates. Digital outputs on the USB-TC unit were in turn connected to the mosfet rectifiers, which switched the power supply to the plates on and off. In this way the temperature at the plates was to be maintained at a set point assigned in the software on the PC. A third thermocouple was used to record the ambient temperature in order to help accommodate calibration of the system. Rain and wind were simulated using a dropper and a spray atomiser for rain while an air pump and a hot air paint stripper were used to simulate wind. In one sample I used ice as a rain source although the data from this may be inconclusive due to difficulties measuring the quantity of ice used. Visual estimation was the only way to measure it and trying to find a 1cm3. It should also be mentioned that neither the dropper nor the atomiser represent very accurate measurement sources with errors I estimate of about  ±0. 5ml. A Velocicalc handheld anemometer was used to measure the wind speeds produced by the wind sources. 3. 2Software Initial work was done on existing VIs (Virtual Instruments) that came packaged with the USB-TC device. These included simple programs to configure the digital ports on the device and to take a reading from an analogue thermocouple port. These helped me understand the workings of Labview and in particular the objects that would be pertinent to my project. The VI developed for the wind and rain sensors in Labview worked on a timed loop basis. The outer loop was timed at the same frequency as the analogue thermocouple inputs to avoid duplicating readings in the data log file. Inside this a second loop was used to continuously change the value at the digital output ports. Depending on the output value of the PID controller a further PID analysis tool would output a Boolean value thus switching the plates on or off. This allowed much more frequent changes to the plates on/off states thus allowing more accurate control of the plates. In order to get accurate time reading for the on/off states of the plates I set them to default off in the outside loop, this meant that in each 0. 5 sec loop iteration the maximum amount of time a plate could be active was 0. 4 sec. This meant that while the system was slower to reach its set point temperature it was less prone to overshooting the set point. This was an improvement on the initial VI that was only capable of switching at the same frequency as the analogue ports and was prone to overshoot the set point temperature by about 1.  °C 2 °C. In contrast the system now recorded average temperatures ~ 0. 5 °C below the set point. The software recorded all data to an excel file which was named arbitrarily according to the timestamp on the machine at the sample start time, e. g. hotplate_data_200713021234. xls. These files recorded about 7200 records per hr and were about 0. 5mb in size before any data analysis. Each file contains rows of readings, which contain a timest amp in milliseconds and 3 temperature readings, top plate, bottom plate and ambient room temperature. Figure 3 shows the time taken to evaporate 1ml of ice over a 10-minute period with the data showing a minimum temperature recorded of 70. 005 °C. The unusual low recorded on the bottom plate is due to the sample being taken before the plate had reached its set point temperature and can be ignored. Figure 4 and Figure 5 show the difference between 1ml of water where the plates can maintain 90 °C in a wind free environment and 1ml of water in a windy environment where the maximum rain plate temperature reached was 83 °C with the rain evaporating at ~69 °C. In both cases it takes less then 400 sec to evaporate the water despite the difference in temperatures at which the plates are operating in each case. It took ~300 sec in the wind free environment to evaporate the water and in the second scenario approx 350 sec. This time difference may have been primarily due to the cooling effect of the wind on the heating element of the rain plate. Figure 1 shows the gap between the plates that allowed the wind to act on the back of the rain plate. In this case I did tilt the plates towards the wind source slightly at an angle of ~10 ° to the horizontal so both plates were exposed to the wind source. This will have had an inevitable effect on both plate temperatures. In a real implementation of a system like this the back of the plates would be thermally isolated with some kind of insulating material thus giving more credence to the results and what they imply in terms of weather effects. However the difference in base temperature can be seen clearly in the second part of Figure 5 where the plates operate at a difference of ~5 °C due to the wind. In this instance the measured wind difference between the plates was 1m/s and with a wind temperature of 25 °C. In this case the wind speeds measured using the Velocicalc were 0. 26m/s at the rain plate, 1. 26m/s at the bottom plate and 3. 52m/s at the ambient temperature thermocouple, which was located about 5cm left of the plates. Average82. 5859389. 4717824. 44208 Min70. 00585. 30724. 179 Max90. 24990. 46624. 67 Figure 3 Ice sample 1ml 10 minutes sample Average83. 9734689. 6681423. 80404 Min72. 47588. 76123. 696 Max90. 36990. 44823. 991 Figure 4 Dropper sample 1ml 10 minutes Average84. 644385. 5546425. 30292 min63. 73476. 00623. 51 max90. 28390. 48229. 856 Figure 5 Dropper sample 10 minutes 1ml with wind source It is also interesting to see that the minimum temperature reached in water was 72. 475 °C as compared to 70 °C for ice and 63. 734 °C for water in windy conditions. This shows that maintaining set point against wind speed will be a strong factor in the power drain of a real system. I took several groups of readings over one-hour intervals and with different quantities of water to try and establish the linearity of the evaporation rate of the system. I also wished to compare the different behavior of the plates due to different methods of applying the water. This would correlate to light drizzle and driving rain in terms of weather conditions. To do this I took half of the measurements with a dropper syringe and half with a perfume atomizer. In this case it could be seen that the surface area of the water particles have a distinct effect on the time taken to evaporate them. Although it is also possible that applying the water with the atomizer may have led to some of the samples missing the tray or evaporating before they contacted the tray, the figures do show an apparent difference in effects.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.