Glass Breaking/Cracking Issue
Description of the Problem
In the beginning of the semester, the team was faced with the issue of glass breaking and cracking in the solar box oven. There were two major hypotheses made on this issue: (1) Shrinkage of wood under high temperature caused normal stress on the glass, (2) Shear stress developed between the silicone caulking and the surface of the glass around the edges. With the glass breaking incident at the end of September, the solar oven team was given an opportunity to test the above hypotheses. The first hypothesis seemed to be irrelevant to the incident, since the oven was designed to have a plenty of room between the wood-frame and the glass. However, the team found chunks of silicone caulking filling in the gaps between the frame and the glass. This could have provided the necessary normal stress to break the glass. The second hypothesis was rejected since it is very unlikely that silicone caulking could have caused such major stress on the glass; silicone is a relatively flexible material.
The team as a whole agreed in the beginning of the semester that the following points must be done: (1) study the strain behavior of glass under changing temperature, (2) investigate how glass can break under high temperature, (3) propose some of the ways that this glass breaking issue can be best avoided. Jay (Hyuk) Jeon chose to investigate the "glass breaking issue."
The glass breaking investigation is not yet complete, so the solar oven team can not yet provide a conclusive technical solution for mitigating or eliminating glass cracking. However, from the September incident and the data collected so far, the team is able to provide a simple safety measure to prevent glass failure. When the empty oven is left unattended and is receiving sunlight, the temperature in the oven rises very high and very quickly; therefore, we strongly urge that the people of Nicaragua do not leave their solar ovens in direct sun while empty for very long.
Strain in Glass Experiment
Omega Bond 200
The previous team from last year failed to obtain correct data for strain in glass under changing temperature; they did not use the appropriate adhesive that would not soften under high temperature. This year, an epoxy called OMEGA Bond 200 was used (figure 1). Its physical properties are also attached below (figure 2). The preparation and curing time of using this epoxy totals to about 10 hours. The epoxy comes as a twin package dividing the resin and the catalysts. First, the package must be heated up to about 150 ºF for about an hour. Second, the "green" divider must be removed to mix 100 parts of the resin (black) and 10 parts of the catalysts (white) until the mix is uniform in color-- the package is already in 10:1 ratio for the resin and the catalysts. After the product is properly mixed, it can be applied to the surface. This adhesive only cures under high temperatures. Curing time takes 8 hours if it is put in the oven under 250ºF, and less time if placed under higher temperatures.
Strain Gages/Wheatstone Bridge System
Both on the glass and the thermally neutral material (ceramic), two strain gages are attached in perpendicular directions (figure 3). This double strain gage was used because the glass is not a perfectly isotropic material; therefore, it will have different measurements in each direction. After the strain gages were attached to the glass and the ceramic, three wires had to be soldered onto each strain gage.
The diagram below (figure 4) depicts the general setup of the Wheatstone Bridge System. In order to record the strain measured by the gages correctly, the system must be setup so that it would become unbalanced when the strain is measured from the gages. When the strain is applied on the strain gages, the resistance of the wire change and the bridge system becomes unbalanced. Then, the voltage is read as a measure of this imbalance and is directly related to the change in the resistance of the wires.
There are two strains that are recorded by each gauge on each material. One is the deformation of glass under high temperature and the other is the deformation of the gauge itself. To obtain solely the strain in glass, the strain that is read on the ceramic must be subtracted from the strain measurements of glass. In order to calibrate, the strain in glass under a certain temperature must be subtracted from the strain in ceramic under the same temperature. Figure 5 shows the strain gauge in use and also highlights the thermocouple and ceramic.
Strain in Glass Data Analysis
After all the preparations mentioned above were complete, Jay Jeon performed a total of two trials for the experiment. One pair of strain gages was attached on the inner glass of the oven and another pair was attached to the piece of ceramic. A total of four thermocouples were used to measure the temperature of ambient air, the oven air, the ceramic, and the inner glass. The halogen lamps were turned on for about an hour and a half to let the oven reach the peak temperature. Then, the lamps were turned off and the oven door was opened to let it cool down. Each test was stopped when the oven cooled down near to about the room temperature.
Temperature vs. Time
The following two graphs show the temperature as a function of time for the two tests.
The above plots show that the glass and ceramic cooled down (Test 2) significantly faster than it heated up (Test 1). The piece of ceramic was placed on the black plate, so it reached the highest temperature of all four measurements in the graph. The temperature of glass was about 20^°^C below that of the oven air for both tests, while the ambient air in the lab stayed the same. Since the cool-down of "test 2" was significantly faster than the heating up of "test 1", the strain will also differ between the two tests, as discussed in the following paragraphs.
Strain vs. Temperature of the Glass and Ceramic
The four plots below show the relationship between strain and temperature in the glass and the ceramic for the tests.
Test 1 was done when the oven was already cooled down to the room temperature while test 2 was done following the cool down of test 1. Therefore, test 2 did not start from the room temperature---the glass and ceramic were at 30^°^C when the test began. This explains the difference in the above two ceramic plots. In test 1, the cool down was moderate since the doors were kept shut as the lamps were turned off. This allowed the glass and the ceramic to recover the strain more smoothly, as indicated by the absence of a gap between heating and cooling in test 1. In test 2, however, the cool down was extreme since the doors were opened and the lamps were turned off. With the faster cool-down rate, the glass and the ceramic responded faster in its strain behavior and the end points reached positive values as they reached the room temperature. In test 1, the strains do not cross zero because it never recovered to its room temperature.
The above plots also clearly show a fluctuation in strain in every temperature interval. This is probably due to strains in both glass and the gage itself. The ceramic strain graph does not fluctuate as much because there is no strain in the thermally neutral material (ceramic); the only strain measurement that is read is the strain in gage itself. As shown in figure 6, the glass in test 2 reaches a higher peak temperature and starts with a higher temperature (30^°C) than the glass in test 1(20°^C). In figure 5, it clearly shows on the graph at what point of temperature the door was opened. This appears to correlate with a significant change in the trend of the graph.
Calibrated Data
The two graphs below are the "true" strain in glass after the calibration of subtracting the ceramic strain from the glass strain. The plots show a fairly linear relationship between the strain and temperature of glass. Also, the difference in the slope between the two perpendicular directions (0 and 1) proves that this glass is not an isotropic material---its strain in the 0 direction is larger than that of 1 direction.
//
Future Experiment
If none of the hypotheses proposed in the "Revisiting the Purposes" are the actual causes of the glass breaking of the solar oven, there is another direction to take for investigating the glass breaking issue. It is possible that the uneven heat distribution and the high rate of temperature change in glass may be connected to the breaking issue. It is known that the temperature inside the oven cannot go up anywhere near to the melting point of glass. Technically, as long as the heat distribution is more or less uniform around the glass and the rate of temperature change is moderate, the glass should not break. The glass would just wait until the glass reaches the melting point without ever breaking. This is similar to the glass breaking phenomenon when the extremely hot water is poured onto a glass surface. Glass is a poor conductor; this means when one side of the glass is significantly higher than the other side of the glass, this will cause high stress distribution within the glass and eventually the glass would break. In the solar oven case, the inner glass is placed in between the inside of the oven and the outer glass. The glass is insulated by the silicone caulking around the edges. The diagram below explains a possible way the inner glass could break.
If the temperature under the inner glass rises quickly due to convection of heat and becomes significantly higher in temperature than the space between the outer and the inner glass, the rate of strain in the bottom will be faster than the top. This can cause enough stress to break the glass if the difference between the top and the bottom of the inner glass is big enough.
This is only a speculation, but it is worth the investigation since it is a common mode of glass failure. In order to investigate this, one would have to measure the temperature on the top and bottom of the glass with the thermocouple while measuring the strain on the top and bottom of the inner glass. From this data, one would be able to see the difference in the top and bottom stress of the glass. This experiment does not require additional or new equipment to test; therefore, the team next semester should consider looking into this mode of failure and investigate whether or not this is a probable cause of glass breaking.
Critical Theory Issues and Challenges
By investigating glass strains and wood shrinkage and expansion, the team is making an effort to find the real source(s) of glass cracking and breaking. If this problem can be solved, the people of Nicaragua can avoid a big hassle of replacing such delicate and expensive material each time it breaks. However, there may be a conflict between our intention to eliminate further glass breaking problem and Nicaraguan's argument that glass breaking is no longer a problem. While we argue that glass breaking problem still exist and we think that it could cause serious injuries, the people in Nicaragua may not believe that glass cracking is a serious issue. Tim Bond has informed us that this is in fact a prevailing attitude . This resistance should be kept in mind when developing any solution that the team might propose to the people of Nicaragua. For instance, the team might want to avoid suggesting any idea that involves a drastic change to the design of the solar oven. The people of Nicaragua might be resistant to adopting such a change since they don't believe there is a problem; however, if the change is less drastic, the people might be more willing to adopt the revision. The solar oven team could also take the approach of striving to explain to the people of Nicaragua why the glass breaking is an issue. While this can be quite a technical subject, it is reasonable to believe that it could be effectively communicated.