Air Conditioning Laboratory Project 99. 10

Appendix

B – file – Appendix B_Team10.xls………………………….43

D…………………………………………...………………...44

E……………………………………………………………...45

F……………………………………………………………...47

G……………………………………..………………………69

H……………………………………………………………..77

J – file – Team10.vi………………….………………………80

K – file – Appendix K_Team10.xls…………………………81

The University of Delaware, Department of Mechanical Engineering, is drastically changing the curriculum for the class of 2000 in order to keep up with ABET standards. Part of this new curriculum includes a joint laboratory class; this lab is for Thermodynamics and Heat Transfer. The professors of these courses, Dr. Wexler and Dr. Advani, have proposed a project to the New Castle Design Associates to design a set of experiments for this joint lab. The problem is to design a set of thermodynamic and heat transfer experiments using a window air conditioner unit for the Undergraduate Laboratory. Experiments will be based on lessons designed by our senior design group.

In evaluating our concepts, we compared them in terms of their correlation to the wants and constraints set forth by our customers. In order to determine how exactly the concepts correspond to the wants we used metrics and target values derived from benchmarking.

Our customer list encompasses our sponsors as well as experts in the fields of HVAC, education, and scholarly work. Our customers are as follows:

1. 
   Dr. Wexler, Professor
   

2. 
   Undergraduate students
   
3. 
   Graduate student TA’s
   
4. 
   Judy Greene, Educational Expert
   
5. 
   Dr. Sun, Professor, Lab Expert
   
6. 
   William Davison, HVAC Engineer
   
7. 
   Manufacturers: Hampden, Armfield Ltd.
   
8. 
   Other schools (Georgia Institute of Technology, Pennsylvania State University)
   

Our list of wants encompasses the top eleven most important wants set forth by our customers. It is as follows:

1. 
   Educational Effectiveness
   

1. 
   Portable
   
2. 
   Easy to use
   
3. 
   Self Evident
   
4. 
   Easy to Set-Up
   
5. 
   Cost Effectiveness
   
6. 
   Multiple Purposes
   
7. 
   LabVIEW Compatible Data Acquisition
   
8. 
   Forgiving of Incompetence
   
9. 
   Quiet
   
10. 
    Quick
    

The constraints for this project include a number of issues. The most important is the size. It must be small enough to be portable because it may be moved often. Also, we are on a strict schedule which states that this project be finished by April 1999, and the format of our presentations and written reports are given, which is all dictated by New Castle Design Associates. Lastly, our budget is four thousand dollars (see Appendix C).

Our mission is to be fulfilled in such a way as to include as many customer wants and constraints as possible. We are going to measure whether we have satisfied this by using metrics and set target values.
Metrics & Target Values:

Although our project integrates two parts, lab experiments and the hardware, some of the wants pertain to the experiments and some to the lab apparatus. Considering the aspect of our lab experiments, each needs to be a self-evident, easy to use, effective learning tool. The most important benefit is that the undergraduate students here at the University of Delaware will learn about heat transfer and thermodynamics and the realities of measured versus theoretical quantities.

To measure how well we have achieved each want, we will use metrics and the target values associated with them. As described previously, our wants are prioritized by their relative importance. Each want has at least one metric associated with it, and some have multiple metrics. As for priority among the metrics, we used the association of the want as a determining factor. For example, if the want is ranked number one the metric corresponding with it is also ranked number one. If a want has multiple metrics, then we considered each to have the same rank. For instance, if the number two want has three metrics, then all three metrics have a rank of two. To see an all–encompassing table of prioritized wants, metrics and target values, see Appendix B. Below we discuss each metric according to whether it pertains to the lab experiments or the hardware.

Several of the wants correspond to the lab experiments; they are educational effectiveness, self evident, and quick. Educational effectiveness is going to be measured by the evaluation chart found in Appendix C, for now, but ultimately by survey during the testing. Whether it is self-evident is going to be determined by survey. The questions on the survey are the basis for Appendix C. At this point in time, Judy Greene has evaluated our lab experiments and commented on them. She used her expertise based on her professional experience to perform the evaluation. When the apparatus is complete, we will persuade some undergraduates to perform the experiments and fill out surveys, commenting on how clear the lessons and objectives are and if the procedure is in a logical order. These surveys will have rankings on them from one to five, with five being the best. We have a target value of 4 because we want it to be self evident to everyone, but we realize that nothing will please everyone. For now we have evaluated these things based on our own experience, as we consider ourselves as experts in the field of laboratories for undergraduate students. Lastly, we want the experiments to be quick. From our personal experiences, those of other undergraduates, those of present teaching assistants, and those of Dr. Advani and Dr. Wexler: if the lab is too long very little is retained. We are going to measure this with time. We have a constraint of two hours, which is the scheduled amount of time for the laboratory. However, we are aiming for no longer than an hour because based on our experience and under the advisement of Judy Greene, students attention deteriorates rapidly at this point.

The rest of the wants pertain to the hardware aspect of the project. The most important according to our SSD chart, as described above, is that it be portable. We are going to measure this by size. The apparatus should fit on a cart that has wheels and is able to pass through a regular doorway, so we have set our target values to be 36 inches in width and 60 inches in height. We are also going to measure this by weight; we have designated 100lbs to be the target value in order to ensure that any one could push the cart.

The next want is easy to use. We are going to measure this by physically looking at the apparatus and determining whether the components are visible and how accessible they are. By accessible, we mean that the students should be able to touch all the parts, not take them apart. In other words, we want to not only see the components in the apparatus, but we want to be able to touch them. We could physically count all the visible parts and the number of touchable parts; however, each apparatus may have a different number of parts so these numbers may not be comparable. Instead, as we consider each apparatus, we will visually examine them and determine whether the components are in fact visible and touchable. Each will receive either a ‘yes’, if they are, or a ‘no’, if they are not, for each category, visible components and accessible components. Our target value is of course yes, because we want everyone to be able to use the apparatus.

We also want the apparatus to be easy to set up. Time is the key method to measure this; we have a target value of 15 minutes, as this is the amount of time between classes and the TA should not need more time than this to set up the lab. We are also going to measure this by the number of separate parts. The idea is to be able to get the entire apparatus onto one moveable cart, because this would greatly minimize the time it would require to set up therefore increasing the ease of setting up the apparatus. Either it fits onto one cart or it does not. Our target value for this is that it does.

The next want regarding the hardware is cost effectiveness. The apparatus should have a low initial cost, a low operating cost, be long lasting, and the parts should be easily available from a store or catalogue. Our target value for a low initial cost is the $4000 budget that we were given. For low operating cost we are aiming for $22, this comes from estimating that the apparatus runs for 12 hours a day, 5 days a week, for four weeks at a rate of 9 cents per kilowatt /hour (This is the rate charges by the City of Newark). For long lasting, we are targeting 10 years because of changes in technology and EPA standards. Currently R-22, the refrigerant in the window air conditioner, is banned from use in car air conditioners and central ac units in homes because of the environmental effects. Right now it is still being used for window air conditioners because of cost considerations, but in a few years it may be illegal to use R-22.

Another want is that the apparatus have multiple purposes. It must contain both thermodynamic and heat transfer principles. This is a yes or no question asked of every concept. Our target value is going to be yes because this is a lab for both heat transfer and thermodynamics so both topics should be included.

This system should also have a LabVIEW compatible data acquisition system so that the existing acquisition board can be used to read the data. This is either present or it is not. Our target value is present. We are investigating other software packages besides LabVIEW; however, LabVIEW is what the ME department currently uses and would be the most convenient. LabVIEW satisfies all metrics associated with the acquisition program required, as it is compatible with itself and it is free.

The apparatus should also be forgiving of incompetence. What this means is that the apparatus should give relatively good data even if the operator makes a few mistakes. This will be measured by physically trying the limits of the apparatus when it is finally assembled, but for now we are taking this into consideration when purchasing parts. We are consulting our customers, such as Dr. Sun, Dr. Advani, Dr. Wexler (who has written a book on sensors and can be considered an expert in this area) and other manufactures of sensors, such as Omega and Optrand. When we are able to test the apparatus, we will compare the data collected, and we have set a target value of 10%. By this, we mean that as long as the error in the data is less than 10%, it is acceptable.

Lastly, we want the apparatus to be quiet. This will be measured by a target value we have set at 60 decibels, which is a conversation level of noise. OSHA laws are not important here, the noise level limitation is for the convenience of the users.

Although the project consists of two integrated parts, the hardware and the lab experiments were benchmarked separately. The following sections, system and functional benchmarking, pertain to the hardware, and the benchmarking for the lab experiments follows below that.

System level benchmarking began with a search for existing thermodynamics and heat transfer labs. We started by looking at what is being offered and what has been offered previously here at the University of Delaware. We found that the MEEG 391, Engineering Science Lab, taught by Dr. Sun, was the closest match, but none of the apparatuses displayed the Heat Transfer and Thermodynamics principles we had in mind. Investigations into various other universities’ engineering departments revealed information of little use. The same result applies to research into HVAC programs in technical schools. Through inquiries to professors and searches on the Internet, manufacturers in this area were found to be Armfield Limited and Hampden Engineering Corporation. Dr. Prasad provided some literature on both the Armfield and Hampden labs. A further check into the companies’ respective Internet pages provided updated information on the old labs and listings of new laboratory offerings.

Armfield Limited offers a large selection of engineering teaching research equipment. The laboratory equipment most applicable to this project can be found in Armfield’s Heat Transfer line of laboratory equipment. Of primary interest is the HT10X line of heat transfer teaching equipment. This line is comprised of single bench top Service Unit, and seven individual laboratory accessories, each illustrating a single heat transfer fundamental. Each of the individual laboratory setups can be purchased with a Data Logging Accessory package. Additionally, in Armfield’s Fluid Mechanics line, the FM23 Plunger Pump Demonstration Unit can be used to partially simulate the compressor found within an air conditioning unit. The air conditioning unit compressor is typically a piston type compressor and the FM23 uses a positive displacement pump. However, this is not a true match. This can also be said for the rest of Armfield labs. Some of Armfield’s products display heat transfer and thermodynamic principles, but none of them do everything a window air conditioner can do.

Hampden differs from Armfield by offering laboratory equipment related to specific industry equipment such as cooling towers, heat exchangers, heat pumps, and refrigeration cycles. Of interest is Hampden’s H-ACD-2 Basic refrigeration Cycle Trainer. According to Hampden’s product literature, this equipment “has been designed to demonstrate the principles of … R-134 a refrigeration system … including heating, cooling, humidification, de-humidification, recirculation, and mixing.” The inclusion of humidification, de-humidification, recirculation, and mixing is beyond the original scope of our project, as this equipment is more oriented to larger, building HVAC systems rather than the small, window unit style air conditioning unit we are interested in. This is also evident in the size of the H-ACD-2. The unit measures 88” x 70” and 31” deep. Total weight for the unit is 1170 lbs. The H-ACD-2 requires 208V, three phase power and a water supply. Hampden offers an option data logging package including 13 type-T thermocouples, a single air velocity meter, three pressure transducers, two differential pressure transducers, and two wattmeters.

Hampden also manufactures a Refrigeration Cycle Trainer, H-RST-2, that covers the complete refrigeration cycle. This does not take the form of a recognizable piece of machinery. All of the equipment, such as the coils, are specially fabricated for this lab apparatus and are not industry standard equipment. The unit is also similar in size and weight as the H-ACD-2. There remains a number of Hampden labs such as the H-RST-3B Basic Refrigeration Cycle and Heat Pump Trainer, the H-6710-CDL Refrigeration Demonstrator, and the H-6830 Heat Pump Trainer that each satisfy a few of our requirements, but are inferior to the H-RST-2 and H-ACD-2 as a competitor.

Overall, we did not learn very much from the system benchmarking. What we did learn was that the window air conditioner appears to be our best choice.

For functional level benchmarking, we researched the major manufacturers of laboratory equipment and scientific sensors. Omega Engineering, Incorporated was found to supply an extensive line of sensors. Additionally, their catalogs and web site offered plenty of technical information about each sensor. The sensors in Omega’s product line that we were interested in included thermocouples, pressure transducers, a mass flow sensor, and a relative humidity sensor. This was beneficial as a source to help determine specific types of sensors need for the purpose of this project.

In terms of thermocouples, we found that type K are the best choice for our needs, the type K range from –100 ^oC to 400 ⁰C. As for the other sensors we are investigating, all of them have outputs of volts, so they will not require modules to convert their signals. Pressure sensors come in a wide variety with fairly large ranges, so we will not have any problems choosing one that will be compatible with the air conditioner. The air conditioner, while off has a pressure of approximately 150psi, and while operating at full capacity, a low pressure of about 75psi and a high of about 250psi. Considering this, we are investigating pressure sensors that have a range of 0 to 500psi. Relative humidity sensors are fairly standard. They all measure 0% to 100% relative humidity. These sensors also determine the temperature of the air, but the velocity will have to be determined with a separate sensor. The mass flow sensors are very delicate. It is difficult to find ones that are compatible with refrigerant. Therefore, the engineering department of Omega is guiding us. We trust their advisement because if it doesn’t work, they’ll replace it.

Another company, Optrand, produces fiber-optic pressure transducers. This provides an alternative method of pressure measurements for the lab as well as allowing a thorough study of a compressor. Optrand’s AutoPSI-TC Dynamic Pressure Sensor would allow continuous measurements of the pressure inside the compressor. This is a new discovery and its feasibility for this particular lab will be further evaluated.

Window air conditioner manufacturers were also researched. A large number of air conditioners in the range of products we are interested in were found in Carrier’s room Air Conditioners series, Airtemp’s Room Air Conditioner series, and Port-a-Cool’s product line to name just a few. The University of Delaware’s HVAC department donated an older air conditioning system. Unfortunately, this system uses R-22 for a refrigerant. We desire a newer system with an environmentally safe refrigerant. It does not do any good for the students to study a system using an obsolete refrigerant. Our customer, Dr. Wexler, agrees with the importance of using an environmentally safe refrigerant for the lab. This air conditioner did, however, provide an excellent opportunity to study the air conditioning system hardware in order to get a better picture of what the eventual lab setup will require.

The LabVIEW software and data acquisition board and computer are already available in the department. Therefore, the selection and purchasing of the data acquisition card and the software does not need to be considered. The number of channels and type of inputs that the board can handle has been determined to aid in the selection of the individual sensors.

Finally, the actual mounting of the sensors needs to be determined. The fittings required in mating the sensors to the air conditioning unit refrigerant lines and coils is dependent on the sensor and its location on the air conditioner. Various tee fittings can be purchased with the required thread to match any threaded sensor and thermowells can be used for temperature measurements. The tee fittings can be soldered to the refrigerant lines, but the sensors should be removable for replacement or maintenance. The wiring and wire connectors required will also be determined later by sensor type and location.
Lab Experiments Benchmarking:

Benchmarking of thermodynamics and heat transfer labs from other schools turned out to be not that beneficial. Of all the schools we researched, none of them use a window air conditioner as a lab apparatus. The most helpful site was that of Michigan State. We got ideas for lab experiment styles and how to prove specific principles from labs with similar heat transfer and thermodynamic principles.

The educational benchmarking we did was primarily through Judy Greene, an educational expert. As a way of measuring the teaching effectiveness of our labs, she recommended we evaluate the lab apparatus and lab reports. She provided us with pamphlets on teaching techniques and writing clear lesson plans. The pamphlets she provided were “Evaluate Your Instructional Effectiveness”, ”Develop a Unit of Instruction”, and “Establish Student Performance Criteria”; we used these to help write the lab manual. She said we would have to evaluate the manual and apparatus for teaching effectiveness with actual students, and the best way to do this would be to conduct a survey and make modifications. We also researched a couple of web sites and books concerning educational effectiveness, however they were not nearly as beneficial as Judy Greene.
Concept Generation:

As previously stated, our project consists of two parts that must be integrated. One aspect is to design the experiments, and the other is to investigate the hardware needed to execute the experiments.

The lab experiments section can also be examined in three different ways. The first being the number of experiments performed during the semester, the second being the basis of the experiments, either by component or by heat transfer/thermodynamic principle, and the third being the way in which the students write their lab reports. The three wants mentioned above for lab experiment concepts are educational effectiveness, self evident, and quick. The key method used to evaluate these concepts is through metrics for educational effectiveness.

For number of experiments preformed during the semester, there are three possible concepts. The first is to do one experiment repeatedly, the only change being that different sensors will be used each time. For instance, we may only be able to evaluate conservation of energy in our apparatus. Conservation would be measured repeatedly, every week, using a different type of sensor each week. This rated very poorly with undergraduates, graduate students, and professors. It rated poorly in terms of educational effectiveness because it would be monotonous since the objectives will be the same every time. Talking with Dr. Greene revealed that as a result, this monotony will cause the students to lose interest. It rated poorly in terms of the number of concepts covered because only a small number of objectives can be achieved, so this set of lab experiments will only cover the principles present. Although, seeing the different answers from different sensors can be an effective learning tool, it is not in our case because the main point demonstrated is experimental error, and these students, being in their third year and having performed numerous labs before, already know what experimental error is.

The second possibility for number of experiments performed during the semester is to do four to seven laboratory experiments. Each experiment would differ in that they would present different fundamentals. This set of experiments would generate more interest because the experiments would vary from lab to lab.

The last possibility is to do 13 experiments. This comes down to one lab per week. Within the apparatus there are a limited number of fundamentals. In order to stretch the fundamentals over 13 labs, the experiments would be short. From our own personal experience, we thought that this would not keep the students interest because the labs would seem pointless.

When considering the basis of experiments, as mentioned above, we considered two options. One is to divide the experiments by fundamentals. Here each lab would deal with a different thermodynamic or heat transfer principle. For example, if the topic is convection, the students would analyze the apparatus everywhere convection is taking place. The overview of how this set of experiments would flow is as follows:

I Conservation of Energy

II Convection

III Conduction

IV Psychometric Chart

V Thermodynamic Diagrams (T-s & P-v)

VI Cross Flow Heat Exchanger

VII Overall Efficiency

This concept contains the fundamentals covered in the classes. With each topic, there would be numerous components and data to analyze.

The other way to divide the experiments is by component within the apparatus. Each lab would deal with a specific part, analyzing every thermodynamic and heat transfer aspects. For example, if the part is a throttle, the students would measure temperature and pressure and show that enthalpy remains constant throughout. Also, the fluid is going through a phase change as it passes through the throttle, so another objective would be to find the quality of the fluid. Another objective would be to determine whether the engineering assumption of approximating the fluid as saturated is appropriate. The overview of how this set of experiments would flow is as follows:

I Heat Exchanger

• 
  Heat Transfer
  
• 
  Efficiency
  

• 
  Conservation of Energy
  

• 
  Constant Enthalpy
  
• 
  Quality
  

• 
  Coefficient of Performance
  

In the third examination of the lab experiments, we looked at the way in which the students write their lab reports. Two options were examined. One is to write standard lab reports after each experiment. By standard we mean they would have all the section including an objective, a procedure, a list of equipment, a theory, a background, results, data analysis, and a conclusion. These reports are typically 5-8 pages. The students would be writing a thorough report on the experiment.

The other option is to write short weekly lab reports; then, in the end, write a technical paper on how the apparatus works. The short weekly lab reports would consist of a short background, results, data analysis, and a conclusion. The technical paper would tie all the labs together and the student would have to support all their conclusions with data that they collected from the labs. One constraint that goes along with this concept is time. The students would need time to write the technical paper after all the labs had been completed and all the short lab reports were written. Our target value is to allow the students four weeks to write the technical paper. The technical paper allows the student to physically understand what is happening inside the apparatus, which gives a complete real world connection.

The hardware section can be examined in two ways: apparatus and sensors. Through conversations with our sponsors, Dr. Wexler and Dr. Advani, as well as a thorough evaluation of our metrics, (see Appendix C) the window air conditioner was selected as the lab apparatus. Most importantly, a window air conditioner is a real world machine that students have seen. This is a point of major concern to our sponsors. The cost of creating a special built apparatus will be well over the cost of purchasing a commercially available air conditioning unit. Finally, the time constraint imposed on us by the senior design process prevent the thorough design of a specialized apparatus in conjunction with the thorough development of the educational values and laboratory procedures. Therefore, the option of creating our own lab apparatus is not feasible. The air conditioner chosen should use the newer, environmentally friendly refrigerant. The older refrigerants are obsolete, so there is no reason to have the students studying an already obsolete system.

Purchasing and instrumenting a window air conditioner is also superior to purchasing a commercially available lab apparatus from our competitors. Again, through discussion with our sponsors and a thorough evaluation of the wants that appear on the evaluation chart, found in Appendix C, the window air conditioner is found to be superior to any commercially available lab apparatus.

The sensor package requirements are partially determined by the lab experiments. The number of sensors required is determined by the data required to complete the labs. Eliminating overlapping or redundant sensors can optimize the cost of the sensor package. The specific sensor concepts related to each individual lab concept can be summarized as follows; instrumented with redundancy, fully instrumented, and economically instrumented.

In the case of the fully instrumented with redundancy, there are redundant sensors for the measurements required, allowing the comparison for different types of sensors. This concept requires the largest and most expensive sensor package. The second case, fully instrumented without redundancy, allows the complete measurement of all points in the air conditioning unit required to complete the labs without the redundancy in the previous case. This does not require a sensor package as large as in the previous case; therefore, it is not as expensive. The trade off is the redundancy factor, which is not a want; therefore, it is not nearly as important according to our wants as cost effectiveness. The last option is to economically instrument the air conditioner. In this case, we would cut the number of sensors to a bare minimum, including just enough to demonstrate a few heat transfer and thermodynamic principles. This would limit the capability of the air conditioner to perform all the requirements necessary to satisfy the want of educational effectiveness. The number of principles will not score well enough on our survey to meet the target value of four. This sensor package is the smallest and least expensive of all the cases, but it trades educational effectiveness for cost effectiveness.
Concept Selection:

In order to propose a complete solution, we need to choose a sensor package, the number of experiments to be preformed during the semester, the basis of the experiments, and the way that the students will write their lab reports. We have already justified using a window air conditioner for the lab apparatus. To completely analyze all of our choices as they compare to our wants, we have created a table that correlate all of our wants with our metrics for each possibility. The metrics are answered with a ‘yes’ or ‘no’, either the metric is met or it is not. This table can be found in Appendix C, the evaluation chart. Once the table is filled out, the percentage of ‘yes’s’ is multiplied by that particular want’s Rate of Importance, determined by the SSD process, Appendix A. Each competing option with the highest number in the end is our best choice. When this is complete, we will have a complete solution.

Upon detailed and careful observation of our metrics on the evaluation chart, we will compare the three cases possible for the number of experiments preformed during the semester. Educational effectiveness is the key want as mentioned above. The second case, four to seven experiments, keeps the students interest better than the other two concepts because it has variety and is complex enough to hold their attention. The first case, one experiment, is too monotonous. It lacks a number of fundamentals and doesn’t keep the students interest. The last case, 13 experiments, is too simple. It doesn’t keep the students interests either because the labs will be too short for the objective to seem clear, as the students would have to come several times to collect the necessary data to perform the analysis. Case two, four to seven experiments, contains more fundamentals than case one, and integrates the fundamentals better than case three. It is apparent using our metrics as they apply to our choices that four to seven labs is the best option for educational effectiveness.

Comparing the two ways to base the experiments, by component or by principle, using educational effectiveness, we see that the basis of component has more real world connections because they are analyzing components. These are things that the students can understand and relate to. The basis of heat transfer/thermodynamic principle analyzes the apparatus by topic; however this does not make real world connections to the students because they are not grasping how each individual component works. The students are only studying fundamentals, and not how they relate to the real world. Knowing that a real world connection is a metric for educational effectiveness, this comparison leads us to the conclusion that basing the experiments on components is the better choice.

Evaluating the choices for the way the students could write their lab reports, leads us to choose the short lab reports followed by a technical paper. By examining Appendix C, it is apparent that this option scored twice as well as the option to write standard weekly reports. The technical paper brings all the fundamentals from the experiments together, forces the students to prove the heat transfer and thermodynamic principles through theory and experimental data, and goes that extra distance to make the real world connection.

Regarding the sensors, the fully instrumented without redundancy was selected. While the fully instrumented with redundancy is more detailed, the additional cost of the sensors prevents its use. The economically instrumented package was not selected due to its hindering effect on the educational effectiveness of our labs. We have estimated the numbers of sensors needed to run our best lab experiments based on components. The specifics are outlined in our Drawing Package, Appendix G. The properties of the sensor are detailed in the fabrication section, as they depend on the specific operation conditions of the chosen window air conditioner.

From the above analysis, it is obvious that our current complete solution is to use a window air conditioner as the apparatus. Approximately four experiments will be written, each in reference to a specific component within the air conditioner. The lab reports will be short and followed by a technical paper, which will summarize the complete function of a window air conditioner, using experimental data to support all statements and analyses.

Fabrication:

Top Cover:

The original air conditioner cover needed to be removed to make the internal components visible during lab. The removal of this cover also exposed the main electrical wiring for the control panel and the easily damaged styrofoam ducts. Finally, the original top cover provided the structural restraint for the cooling coil and fan. A new cover needed to be fabricated that would protect the foam and wiring, retain the original structural integrity, and keep the major components visible and accessible. A new top cover was designed and galvanized steel selected for its durability and ease of fabrication. (See Figure 1, Appendix G)

Fabrication of the top cover began with the initial laying out of the pattern onto a sheet of galvanized steel. Next, 0.125” diameter holes were drilled at all internal corners. These holes provided a point to cut to as well as allowed clean bends to the corners. The pattern was then sheared from the sheet with a jump shear and finished with aviation snips. All bends were made on a brake. The cover is installed with self-tapping sheet metal screws, therefore no mounting holes are provided. The cover was left unfinished, bare galvanized steel.

Three Safety Guards are required for the apparatus. (See Figures 2, 6, & 5, in Appendix G) The air conditioner uses two fans, which were exposed by the removal of the original cover. A guard must be placed over the opening between these fans to prevent injury due to the exposed fan blades. Additionally, the fragile coil fins were exposed with the removal of the original cover. These fins need to be protected from damage. The fins are also sharp and therefore cannot be left exposed. A galvanized steel frame with galvanized steel screen was selected for ease of fabrication and availability. The screen is ¼” mesh to provide sufficient protection and maintain visibility of the protected components.

The first step in fabricating the guards was the laying out of the frame onto the sheet steel. The steel guards were cut on a shear and bent on a brake. The screen was cut with aviation snips. The frame was installed on the screen using 1/8” pop rivets. The guards were finished with black wrinkle-finish paint. The guards are installed with self-tapping sheet metal screws, therefore no mounting holes are provided. Heavier gage steel is recommended for future guards. The heavier steel would allow for cleaner bends and a stronger guard.

Thermocouple and Pressure Transducer Connector Panels:

One of the wants for the lab apparatus was portability. To increase the ease in which the apparatus can be connected to the computer and to clean up the wiring, panels were created to mount connectors for all of the thermocouples and pressure transducers. Two panels were designed, one for thermocouples and one for the pressure transducers. Each panel is similar in design, varying only through the dimensions of the connector holes. Two separate panels were selected rather than a single panel to accommodate future changes or upgrades in either the thermocouples or pressure transducers. (See Figure 4 in Appendix 4)

The panels were cut from .080” aluminum sheet on a band saw. 0.125” holes were drilled at all internal corners before cutting. The connector mounting holes were fabricated by drilling holes for each corner then cutting and filing the rectangular hole. For future production, the purchase and use of a square or rectangular punch would dramatically decrease the time required for fabrication as well as result in a much cleaner and accurate panel. All holes were drilled on a drill press after laying out the hole locations with a prick punch and center punch. The additional accuracy of the milling machine was deemed unnecessary for these panels and the drill press was chosen for speed. After all holes were made, the panels were bent on a brake. The guards are installed with self-tapping sheet metal screws, therefore no mounting holes are provided.

The pressure sensors mount with a 1/8”NPT male connections. Bullet piercing valves were used to tap the pressure transducers to the refrigerant lines for two important reasons. First, using the piercing valves eliminates the need to cut the lines and sweat in fittings. Secondly, in the event of a pressure transducer failure or the need to change transducers, the valve can be shut off and the transducer removed without the need for evacuating the system of refrigerant. The valves use a ¼” male flare fitting. A 1/8”NPT to flare adapter was used in conjunction with a swivel nut flare fitting to mate the parts properly. The swivel nut flare adapter was used to allow rotation of the transducer with relation to the valve in order to guarantee proper alignment before tightening the seal. The No. 1 pressure transducer location required a clearance cut in the air conditioner chassis. This cut can be made properly with a die grinder. (See Figure 3, Appendix G)

Wiring of the transducers was done after installation. The transducers share a power source with the relative humidity sensor. All of the transducers’ power and ground leads are tied into a single two-pin connector on the pressure transducer connector panel. A power lead was made with a mating two-pin connector. The outputs of each of the pressure transducers are connected to two-pin connectors, even though the output only uses a single wire. Some pressure transducers and transmitters use a two-wire output. Using a two-pin connector allows the future modification to a different style of pressure transducer without changing the connector arrangement.
Mass flow sensors:

The mass flow sensor is supplied with two compression fittings. The sensor is heavy enough to require support other than a simple hanging from the copper lines. The bottom of the sensor has two tapped holes for 4-40 machine screws. The top cover was removed and the machine screws threaded in from the bottom. University of Delaware HVAC spliced additional copper tubing into the lines to route the refrigerant through the mass flow sensor.

The thermocouples are supplied with 36” leads. The leads were cut 10” from the thermocouple, creating a thermocouple with 10” leads to be mounted on the air conditioner and a 26” long lead to run from the air conditioner to the data acquisition board. A female miniature thermocouple connector was installed on the thermocouple leads while a male miniature thermocouple connector was installed on the data acquisition leads. These connectors are installed by unscrewing the connector halves, carefully stripping the insulation off the leads, installing the leads at the connector’s screw terminals using the soft nylon washers, and reassembling the connector halves. Additionally, the panel mount brackets were installed on the female connectors.

The relative humidity sensor requires an external power source. The sensor uses a four pin connector. The female end of the connector is wired to the sensor, the male end is supplied with the sensor. The power source and output leads need to be soldered to the male connector. The connector is disassembled by removing the two screws holding the tail piece on and the single screw holding the halves together. The wires are soldered to the sockets in the connector and the halves reassembled.