Paper

Transcript

EXPERIMENTAL ANALYSIS OF CRACK DISPLACEMENT ON THE WATTS TOWERS August 21, 2013 Winston Boyce Home Institution: University of California Los Angeles REU Site: University of California Los Angeles Principal Investigator: Dr. Robert Nigbor Abstract The Watts Towers are a national historic monument located in Watts, Los Angeles, built between 1921 and 1955 by Simon Rodia. Overtime, cracks have formed and expanded in the concrete as the towers have decayed. These cracks have been repaired over the years, but continually reopened. As part of a combined LACMA/UCLA effort to preserve the site, displacement sensors were placed on north and south side cracks on the base of the Central Tower to monitor movement of the cracks. Data showed that the north crack daily displacement was inversely proportional to temperature, while the south had a bimodal shape. The hypothesized reason was because the south crack is in direct contact with the sunlight all day while the north crack is largely shaded. Covering the south crack with an 8x12 tarp changed the shape of the south crack, but other factors are affecting it. Introduction South Central, Los Angeles has always been known for its history rooted in violence and crime. The history of the Watts area is a prime example, being mainly known for the Watts riots which were a result of the Rodney King beating video becoming mainstream. The Watts Towers of Simon Rodia happen to be one of the exceptions, however. They are located on 107th street in one of the most gang prevalent areas of Los Angeles. Although there is such a blemish on this area, the towers have proven to be that one ray of sunshine on a cloudy day. The towers are considered a neutral zone for the gangs, and have a sense of peace and serenity around them. Simon Rodia (pictured in Figure 1) was an Italian immigrant who found his niche as a construction worker. He built the towers in his backyard from 1921-1955 as a sort of recreational project. The three main towers all exceed 60 feet in height, with the two tallest reaching 97 and 99 feet. Rodia constructed the towers using normal, everyday tools, pipe fitter pliers, and a window-washer belt and buckle. He didn’t use any machine equipment, scaffolding, bolts, rivets welds, or any other technological advantage that could’ve helped with the construction. Upon completion of the towers, he moved away from Watts and never returned, giving ownership to one of his neighbors. A fire in 1956 burned down his house, but the towers survived. The Los Angeles County Museum of Art (LACMA) has begun a long-term preservation plan and has enlisted the help of structural engineers from the University of California, Los Angeles (UCLA). Figure 1: Rodia climbing the tower to continue his construction Over the course of time, cracks have been forming and expanding in the concrete as the towers have decayed. These cracks have been repaired over the years, but have continually reopened. As part of the combined LACMA/UCLA effort to preserve the site, displacement sensors have been set up. They are located on north and south side cracks on the base of the Central Tower to monitor movement of the cracks. Figures 2 and 3 show the structures and their interconnecting braces, which have helped increase stability. Figures 4 and 5 show the three main structures and how they interact with the other 14 structures. 1 Figure 2: The braces between the structures were put in as structural stability precautions by Rodia after the 1933 Long Beach Earthquake Figure 3: The interconnected nature of the structures allows them to more effectively withstand earthquakes and high speed winds 2 Figure 4: The three tallest towers range from 60-99 feet in height. The East and Central Tower are connected and have a coupled motion while the West Tower moves independently of the other two Figure 5: The towers are not just comprised of the three main towers, but actually include an additional 14, smaller structures 3 Lit Review The Watts Towers consist of 17 interconnected structures that were built over a 30 year period as a hobby for a construction worker, Sabato “Simon” Rodia, in the backyard of his house. They have been a national historic landmark since 1990 and are constructed of steel pipes concrete, and whatever else Rodia could find lying around (Den Arend, 2013). After Rodia sold the property and moved away the city of Los Angeles attempted to raze the towers, but the towers had already garnered national attention. Activists attempted to preserve them, but the city still wanted to have them torn down. In an attempt to justify their actions, the city organized a test of the structural stability of the structures (“Watts Towers,” 2013). The lead engineer on site, Bud Goldstone, administered the test along with architect Edward Farrell. They tested only the tallest tower on the site, subjecting it to a horizontal force of 10,000 lb, the equivalent of the required code wind force, for one minute, at which point the main beam used to subject force on the tower yielded. At the end of the day, the structures were deemed stable and allowed to stand (Goldstone, 1963). The towers’ stability was tested again in the Northridge earthquake of 1994, and repairs had to be made to the towers (“Watts Towers,” 2013). Evaluations have been done on the towers throughout the course of time, but each repair done to the structure has modified it slightly (Pugliesi, 2004). This mindset led to the desire for better modeling of the towers. Questions about how the towers were affected by its environment also began to be raised. In 2011, nine USB accelerometers were placed at elevations of 2.5-4.2 meters (8.2-13.8 feet) on the West Tower, Central Tower, East Tower, Ship, and overheads, and recorded approximately 48 hours of data (Schweri-Dorsch, 2012 ). Two seemingly contradictory conclusions were provided in the report. Vibration data collected during the Santa Ana wind storm of Nov. 30th–Dec. 2nd, 2011 confirms that strong winds cause vibrations and wind strength and direction influences vibration magnitude and direction (Schweri-Dorsch, 2012). Thus Schweri-Dorsch indicates that wind is one of the factors affecting the vibrations of the tower, but the report also says that data collected from monitoring suggests that winds experienced in Santa Ana wind events in South Los Angeles do not cause vibrations great enough to cause structural damage to Watts Towers (Schweri-Dorsch, 2012). Therefore, he also concludes that although the wind may cause alterations in the vibrations, ultimately wind will not damage the structural integrity of the towers. Over the course of time there have been uncertainties about instrumentation used for engineering studies. As a result of these uncertainties, tests on instruments were performed by Skolnik, Nigbor, and Wallace (2009). The results showed that for single-channel accelerationsensitive results, a minimum sample rate of 200 samples per second is sufficient to limit errors to less than 1%. Short period responses are limited to 5% error, and higher sample rates are required to achieve an error of 1% (Skolnik et al., 2009). This information led to the current tests in which three accelerometers, two displacement sensors, two tilt sensors, two temperature sensors, one wind speed and one wind direction sensor are being used to monitor the towers according to these specifications. Previous tests were done to see how temperature affects the Watts Towers. One conclusion garnered was that there is a correlation between temperature, embedded steel strain, and exterior gap change. As the temperature decreased, the steel strain increased and the crack gap initially decreased then increased (Anco Engineers, 1988). In addition to temperature and wind, data has 4 previously been examined to see how other environmental factors, such as rain and earthquakes, affect structures in general. On the Network for Earthquake Engineering Simulation (NEES) model test structure in Garner Valley, California, tests were done to characterize the interaction between the soil foundations and the structure. The NEES model test structure is a one-story soil-foundation structure-interaction (SFSI) structure, located in a seismically active region of California, used for ground motion research. The results from the tests at this site concluded that accounting for foundation/soil changes properly does not affect vibration periods significantly, but significantly impacts the distribution of inter-story drifts over the height of the structure (Tileylioglu, 2008). These results show that although the foundation doesn’t significantly affect vibrations, it does increase the tilting of the structure the higher up you go. The results from all these tests show that earthquakes, wind, and sun affect the towers. This study will focus on determining which condition causes the biggest change, and what techniques can be used to prevent further deterioration of the towers. Background Crack Displacement Data from sensors that were previously placed on the towers shows that the crack displacement changes throughout the day in accordance with the change in temperature. Figures 6 and 7 below show crack displacement graphs for data collected on June 26th. -0.2268 35 -0.227 30 -0.2272 25 -0.2274 20 -0.2276 15 -0.2278 10 -0.228 5 0 Crack Displacement (in) Temperature (Degrees Celsius) June 26th 40 -0.2282 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hours of Day Temperature South Crack Displaacement (in.) Figure 6: The south crack displacement decreases as temperature increases, but then it increases around the peak in temperature, then decreases and increases again 5 June 26th 40 -0.2052 35 -0.2053 -0.2055 25 -0.2056 20 -0.2057 15 -0.2058 -0.2059 10 Crack Displacement (in) Temperature (Degrees Celsius) -0.2054 30 -0.206 5 -0.2061 0 -0.2062 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hours of Day Temperature North Crack Displaacement (in.) Figure 7: The north crack displacement is essentially inversely proportional to the temperature as it decreases as temperature increases, then increases around the peak of the temperature The most probable conclusion for the reason of the displacement reaction to temperature can be that as the concrete around the crack heats up, the concrete will expands and the crack begins to close. Concrete, however, isn’t the only thing affecting the cracks. Reinforcing steel and the global tilt behavior have a part in the changes of the crack displacement as well. As the sun rises and heats up the steel, the side that is facing the sun will expand, forcing the tower to tilt the other way. Over the course of the day, the tilting forms an elliptical pattern directed away from the sun. Throughout the day, as the towers tilt towards the crack, the displacement decreases, but as it tilts away, the displacement increases. The last factor affecting the cracks is the steel itself. As the steel expands, the displacement increases, and as it contracts, the displacement decreases. Therefore the graphs obtained (Figures 6 and 7) are a combination of all these factors. Orientation of the Towers The way the towers are oriented also plays a role in the difference between the crack displacement graphs. This means that the crack located on the north side of the central tower is shaded all day from the rays of the sun. In the morning the East Tower blocks the rays, throughout the course of the day the rays are blocked by the south side of the tower itself, and in the late afternoon/evening the West Tower blocks the rays. Figure 8 displays the shading that the north side crack receives. 6 Figure 8: Crack displacement sensor on the north side crack. The shade from the towers is seen. The south side crack (Figure 9) doesn’t have the same blocking as the north side crack however. Since the sun rises and sets on the southern side of the towers, the south crack is in direct contact with the sunlight from sunrise to late afternoon/evening. This causes more factors to come into play with the south crack displacement in comparison to the north crack. Figure 9: Crack displacement sensor on the south side crack. The direct contact from sunlight is seen. 7 Methods and Equipment One tri-axial accelerometer was used on the Central Tower to see which direction the tower is vibrating. One biaxial tilt meter was used to see how much the tower is swaying throughout the course of the day. There are two temperature sensors, the north and south crack displacement sensors (Figure 10), a voltage sensor to make sure the input into the displacement sensors is stable, and a wind sensor (Figure 11). The wind sensor serves three purposes as it tells not only the wind speed and wind direction, but it also serves as the second temperature sensor. Figure 10: One of the crack displacement sensors recording data Figure 11: The wind sensor is located slightly off of the Central Tower, but captures the measurements of the wind that affects the tower 8 2 Contributions My part in the project was to frame a hypothesis of why the south crack displacement is so drastically different then the north crack displacement throughout the course of the day. In order to do this I had to analyze all the data from January to June 2013 and formulate a hypothesis. Hypothesis Due to the fact that the south crack is in direct contact with the rays from the sun for the majority of the day, the hypothesized theory was that the sun was the main reason for the difference. As the sun heated up the concrete, it expanded, causing the crack to close. The steel also heated up, but there was a delay in the process due to all of the covering on the steel, but once the process reached the steel, it would expand, causing the crack to reopen. In order to check if this hypothesis could possibly be correct, cloudy days, or days when the sunlight wasn’t prevalent were first checked.This was done using solar radiation data from the LACMA weather station located on-site. Charts were then constructed for average daily radiation (Figure 12) and peak daily radiation (Figure 13). The days for which the daily average value for solar radiation fell below 150 and the daily peak value fell below 500 were then analyzed. Data was then checked for the days over this period where both the average and daily peak were below the specified thresholds because it was possible that these days were cloudy. Average Daily Radiation 450 400 350 Radiation (Dose) 300 250 200 150 100 50 0 0 10 20 30 Days Elapsed 40 50 Average Daily Radiation Figure 12: Average Daily Solar Radiation over 53 day period 9 Peak Daily Radiation 1400 1200 Radiation (Dose) 1000 800 600 400 200 0 0 10 20 30 Days Elapsed 40 50 Peak Daily Radiation Figure 13: Peak Daily Solar Radiation over 53 day period Solar radiation can be a good indicator of an overcast day because when the dose reading is low the amount of radiation being absorbed from the sun is low. This can possibly be an overcast day because if low amounts of radiation are being absorbed by the weather station, then the sun’s rays may be stymied. April 12th was selected for further investigation because it had the lowest average and the lowest daily peak radiation values. Figure 14 shows the temperature and crack displacements on that day. April 12th 24.000 1.40E-03 1.20E-03 1.00E-03 16.000 8.00E-04 6.00E-04 12.000 4.00E-04 8.000 2.00E-04 0.00E+00 4.000 -2.00E-04 0.000 -4.00E-04 -4.000 -6.00E-04 Hour of the Day Termperature South Crack Displacement North Crack displacement Figure 14: April 12th Crack Displacement Graph 10 Crack Displacement (In) Temperature (Degrees Celsius) 20.000 The April 12th Crack Displacement graphs corresponded to the lowest point on the “Peak Daily Radiation” and “Average Daily Radiation” graphs. Graphs for the rest of local minima were created as well. Upon creation of graphs for these low solar radiation days, it was noticed that the crack displacement graphs didn’t show change in reference to the difference between the south and north side cracks. A plan to test the hypothesis was then created. Testing The purpose of the experiment was to test the hypothesis that the irregularity in the south side crack displacement was resultant from the crack’s direct exposure to the sunlight. The north crack displacement graph showed an inverse trend between displacement and temperature, which made sense, because the effect of the sun wasn’t really prevalent on the crack. The south side crack, on the other hand, had a bimodal graph. This was unexpected because of the simplicity of the north side crack. By examining the structure itself and the location of the cracks, the hypothesis was formed that the irregularity in the south side crack was from the exposure to the sun’s radiation. In order to maintain structural preservation over time, it must first be answered why the displacement in the crack occurs the way that it does. The plan for the experiment was to set up a small tarp that covered the portion of the Central Tower where the south crack displacement sensor was located. Then after data was collected the north and south crack displacement graphs were compared. For set up of the experiment, we attached the top corners of a tarp (Figure 15) to the “beams” of the Central Tower via ropes so as to leave space between the tarp and the tower so it isn’t in direct contact with the tower, but sufficiently blocks out the sunlight. Then we attached the bottom corners, one to a base column on the West Tower, and the other to the wall to the south that spans the length of all 17 structures (Figure 16). The tarp was then left up for two days. Two days of data is sufficient enough to get a comparison, but the second day was mostly overcast, so we only used the data from the first day. Figure 15: The 8x12 tarp that was used to cover the portion of column where the south side crack was located 11 Figure 16: The process of setting up the tarp and using ropes to attach it at the bases. A great deal of carefulness was taken because we couldn't damage any of the monuments Although crack analysis was the main focus of the experiment, other factors were taken into account. The infrared thermometer was used to measure the temperature inside the shaded tarp at 12 locations on the column every 30 minutes, as shown in Figure 17. We then measured the temperature of a neighboring column on the south side that wasn’t in the shade to compare data between the two. Figure 17: The infrared thermometer that we used to take temperature readings along the front and back side of the column 12 Results The results of the test came back showing that the sun’s rays being in direct contact with the crack was a reason for the difference between the two graphs. The shade experiment gave a change in the graph for south crack displacement, but it was not the change that was expected. Figure 18 shows that there was a difference in the south crack displacement graph before and after the shade experiment, but it is complex. Without the shade experiment, the crack first closed, then opened, then closed again. With the shade up, the crack first opened, then closed, then opened again. 4.00E-04 30 2.00E-04 25 0.00E+00 20 -2.00E-04 15 -4.00E-04 10 -6.00E-04 5 -8.00E-04 0 Crack Displacement (in) Temperature (Degrees Celsius) Shade Experiment 35 -1.00E-03 0 6 Temperature 12 Hour of the day South Crack Displacement 18 24 North Crack Displacement Figure 18: Results of the Two Day Shade Experiment Conclusions Although this information portrays the hypothesis as a significant reason for the difference in the crack displacement graphs, direct exposure to sunlight was not the only factor influencing the crack’s displacement. The experiment covered about eight feet of the column, leaving around 90 feet still exposed to the sun’s rays. Therefore, the experiment only accounted for local effects, but not global ones. Modeling on the towers shows that throughout the course of the day, the towers actually tilt away from the sun in an elliptical pattern. As a result, the global effect of tilt was still prevalent, as we didn’t shade a greater portion of the tower. Future research can be done to better understand the behavior of the cracks and would involve a few changes. Some changes that can be made are using a more reflective tarp, covering a greater surface area so as to eliminate global effects like tilt, and to test for a larger amount of time. A possible hypothesis for future testing would combine the phenomena of tilt with the contact with the sunlight. 13 Contact information If you have any further questions, please contact me at [email protected] or Jackson English at [email protected] or Dr. Robert Nigbor at [email protected]. Acknowledgements This research was partially supported by the National Science Foundation through the Research Experience for Undergraduates program (EEC-1263155) and the George E. Brown Jr. Network for Earthquake Engineering Simulation (NEES) Cooperative Agreement CMMI0927178. Any opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the National Science Foundation. I would like to thank Dr. Robert Nigbor and Jackson English for their mentorship and guidance on this project. I would also like to thank the entire NEES@UCLA staff, Erica Eskes, Alberto Salamanca, Steve Keowen, and Andrey Kozhukhovskiy, for all of their help and support. References Anco Engineers. (1988). “Simon Rodia Towers Environmental Measurements”. Tech. Report. Culver City. Den Arend, L. (2013).” Watts Towers by Sam Rodia”, Retrieved from (August 14, 2013) Goldstone, N. J. (1963). "Structural Test of Hand-Built Tower." Experimental Mechanics, 3(1), 8-13. Pugliesi, Raymond S. (2004) Summary of Investigation. Memorandum. Degenkolb Engineers. Rodia, Sam (2006) “Watts Towers” < http://www.wattstowers.us/history.htm > (Aug, 20,2013) Schweri-Dorsch, Sylvia. (2012). “Watts Towers Conservation Project.” Internal Report. Los Angeles: Los Angeles County Museum of Art, California. Skolnik, Derek A., Robert L. Nigbor, and John W. Wallace. (2009). A Quantitative Basis for Strong-Motion Instrumentation Specifications. MS Thesis, University of California, Los Angeles, California. Tileylioglu, Salih. (2008). "Evaluation of Soil-Structure Interaction Effects from Field Performance Data." Dissertation, Civil Engineering Department, University of California, Los Angeles. “Watts Towers.” (2012). In Wikipedia, The Free Encyclopedia. (July 9, 2013). 14