An Upgraded Control and Data Acquisition System for the Universal Element Tester at the University of Houston

Transcription

1 An Upgraded Control and Data Acquisition System for the Universal Element Tester at the University of Houston Brent Lindelof, REU Student A. Laskar, Graduate Mentor And Y. L. Mo, Faculty Member Final Report Department of Civil and Environmental Engineering University of Houston Houston, Texas Sponsored by National Science Foundation August of 12

2 Abstract A universal element tester (UET) was constructed in at the University of Houston to perform biaxial or triaxial tests on full-size reinforced concrete panels 55 in. square and up to 16 in. thick. Each panel element represents an element cut out from large structures, such as bridge girders, shell roofs, nuclear containment structures, concrete offshore platforms or high-rise shear wall. Despite previous upgrades to the control system, the UET still has one serious weakness: The loading procedure has to be programmed before tests and it cannot be changed during the tests. This limitation is undesirable for cyclic loading. The complex load path of such loading often requires changes during the test (run-time). To add this capability, it is necessary to develop a new electronic control system with custom-made computer software. This paper presents the requirements for upgrading the equipment and discusses the enhancement of flexibility of operation and breadth of application. Introduction The University of Houston constructed a universal element tester (UET) in 1988 with help from grants through the National Science Foundation. It was built in the Thomas T. C. Hsu Structural Laboratory. Structures like box girders, bridges, shear walls, and offshore structures can be visualized as an array of membranes elements. These elements can be constructed and tested in full-scale in the UET. It has the ability to test reinforced concrete panels 55 inches by 55 inches and up to 16 inches thick. It is considered to be full-scale because it can handle steel reinforcement up to 1 inch in diameter. The UET has been used to develop the constitutive laws of reinforced concrete subjected to biaxial in-plane loading. These studies have revealed that softened compressive stress-strain curve of concrete is a function of four variables: (1) the loading path of the two principal stresses, (2) orientation of the steel bars with respect to the direction of principal stresses, (3) the ratio of steel percentages in the longitudinal and transverse directions, and (4) the compressive strength of concrete. Once these variables are defined, the behavior of many different structures can be accurately predicted. However, the UET has serious drawbacks for the advancement of research. To continue developing the constitutive laws of more complex concrete designs, the software and hardware needs to up-graded to accommodate new panel designs and more dynamic loading conditions. With current technology, this system can be utilized to its fullest potential, yielding the desired results for seismic activity on RC structural membranes. Once this is achieved, the UET will truly be state-of-the-art technology for the advancement of structural mechanics. 2 of 12

3 Universal Element Tester General Description As mentioned earlier, the UET was designed to test concrete panels 55 in. by 55 in. and up to 16 inches thick. Refer to Figures 1 and 2 for north and south views of the UET. The test element (panel) fits inside the reaction frame along with 37 in-plane jacks and 3 rigid links. There are also 3 out-of-plane rigid links attached to the panel. The links are attached to satisfy static equilibrium requirements, and are matched to the same strength and length of their corresponding hydraulic jack. Each jack is capable of producing 250 kips at 5000 psi and are equipped with spherical hinges at both ends to control the alignment of the applied forces. The loads are applied to the panel at five equally spaced locations along the four edges of the panel. The jacks are attached to the panel through connector yokes and the load is applied through a sophisticated hydraulic system, consisting of a 5000 psi pump unit and a complex series of valves and hoses. The current hydraulic and control system is illustrated in Figure 3. Control of the jacks can be executed either by servo-control system or by a set of 60 manual controls. The servo control system has ten channels, each with a servo controller and a servo-valve package. Each package consists of a servo-valve, a manifold with ten pairs of outlets, and a pair of Delta P pressure transducers. Each channel provides a different oil pressure in the manifold. Any of the ten different pressures can be supplied to any jack through a pair of flexible hoses with Quick-disconnect fittings. The hoses are connected to the pair of desired jack terminals, which is connected to the jack through a pair of steel tubes. The ten different oil pressures are channeled to the 37 jacks in a way that satisfies threedimensional equilibrium. The oil pressures in the manifolds are controlled electronically by ten controllers. The closed-loop control of the servo-valves is achieved with information supplied by the pair of Delta P pressure transducers, the linear differential transformers (LVDTs), and the jack load cells. Loading sequence of each controller is programmed by an interface servo-computer to achieve either strain control or load control. This system achieves two main objectives. First, the load application is very versatile because each of the ten oil pressures can be load control or strain control. Second, the load application can be computer controlled, thereby simplifying the load operation of 37 jacks. The oil pressure can also be operated manually by disconnecting the hoses from the manifolds and using the manual levers. This is very important when loading and unloading the panels, so the jacks can be lined up with the yokes for the 90 mm diameter steel keeper pin to fit in the rod-eye hole. Without the manual control, lining up the panel yokes with the hydraulic jacks would be virtually impossible. 3 of 12

4 This servo is a very versatile system giving the programmer the ability to use strain control or load control; however, the mode of control cannot be changed once the program has started and cannot be changed during the test. This is an undesirable limitation for such testing procedures. It would be to the advantage of the users to be able to use load control while the load is increasing and then switch to strain control while the load is decreasing. This real-time control would slow the cracking process of the panel and give the opportunity to collect more data, by which more information could then be drawn about the behavior of the concrete and steel during the portion of the test where the load is decreasing. With the current system, this is impossible. Data Acquisition In order to obtain data during the tests, there are one hundred channels available that run into a Hewlett Packard 3852A (HP 3825A) data acquisition processor. This device collects the data from load and deformation transducers; including load cells on all the jacks, LVDTs on the faces of the panel, and strain gauges on the reinforcing bars. There are 10 LVDTs on each face of the panel; four horizontal, four vertical, and two diagonal. The LVDTs are attached to aluminum brackets, with each direction a different height so they will not touch when crossing paths. The brackets are attached to the panel by all thread that is cast into the panel. This is done so undesirable movements of the brackets due the local cracking near the rod can be avoided. The test area is a 31.5 inch square in the middle of the panel. Refer to Figure 4 for the LVDT arrangement. The load cells are attached directly to the jacks. All of the data is transferred in micro volts from the HP3825A to a PC for processing and storage. There are two computer programs for the data processing and imaging. The first one controls the test, organizes the data, and stores the file. The second one analyzes the collected data and sends the data to another CPU that manages the data.. This data management has 56 signal conditioners, 64 channels of storage, and custom designed software to display the data in real-time graphical display. This system needs to be upgraded to include software and hardware that will enable the user to control the test while it in process, giving more flexibility to the system and types of tests it can simulate. Procedures For the actual test to be conducted there is a tedious procedure that must be followed exactly or the test will fail or abort. The procedure is as follows; first the instrumentation, panel connections, and hydraulic manifold must be checked to make sure everything is attached properly. Then the hydraulic pump is turned on and the computers are booted up. Then the program must be written into the computer. For this, 4 of 12

5 there are many steps. First, a reference file must be created. To do this, LVDT calibration data must be entered into the CPU and the test name and data is entered. Then the control boxes must all be zeroed in the proper order. After this the file is stored as a zero file. The test procedure can then be entered into the CPU using end levels to define the control type (load or strain), magnitude, and length. The first three end levels test the system by bringing the load up to 5 kips over five minutes and then back to zero. This is a test sequence to make sure the system is working properly. After the programming is complete, make sure the printer is properly hooked up with plenty of paper, and begin the test. During the test, it is important to print out the data screen at the end of each end level and to record the crack sizes in the concrete with 30 handheld microscopes with a grading of inch. The cracks area also marked and photographed. However, as with all technology, there are limitations that need to be addressed so that further advancements can be made in this area of research. Drawbacks of Operation The UET is a sophisticated piece of hardware that has been necessary for the development of the constitutive laws of RC elements. To ensure the future of this research, the drawbacks of this system need to be discussed and solved. The primary problems with the UET lie within the computer software and hardware arrangement. The first area that has proven to cause trouble is the fact that any error that occurs during the test will cause the system to immediately abort. Any small error, including input level errors, low oil pressure, power surge, and even printer errors cause the entire test to stop. Once this has happened, the user must reset everything and cannot continue where the test stopped. These small errors can be overcome by simple coding within the software package. The second area that is problematic is the user-interface. This system is not user friendly. The control boxes must be manually zeroed, which is very tedious and time consuming. If the test aborts, the boxes must be re-zeroed. To do this the entire system must be shut down and rebooted. The software needs to have user defined data channels and data, fields. When these are predefined for default, there is no flexibility the data acquisition. The software has unnecessary elements that are not applicable to this system and need to be eradicated from the system. This system also has no fail-safe for the data and program; when errors occur, the latest data is lost, and the system cannot pick up where it left off. The most significant drawback of this system is the lack of real-time programming. There is no current way to change any test parameters once the start button has been pushed. The only thing the user can do interactively is pause the test. The need for real- 5 of 12

6 time load control and strain control is necessary for cyclic loading and data collection. There is no time device for the test in its entirety. There is a clock, but it does not tell you when it starts, or when each end-level begins. There is not an auto-save option for running data. Therefore; if the test fails, due to any one of the minor errors, the current data is lost unless the user preemptively prints the data before failure. This has happened many times to the users of this system, and has proven to make the research slow and difficult. With remediation of these problems, the research will be more versatile, and easier to complete on schedule. Upgrade of Current System The University of Houston is currently upgrading the computer software and hardware for the UET. The upgrade is being completed by Gardner systems, as they helped with the development of the original software and have the expertise necessary for this project. As mentioned, the upgrade consists of new software and hardware for the computer system that controls the system and collects the data. Gardner Systems has rewritten existing DOS-based block program to Windows-based operating system. This new software provides an additional function to modify load profile while a test is running. This eliminates the need to abort tests, re-enter parameters, and continue test. The control and analysis functions are also performed and controlled on the same computer. The upgrade also offers Ramp Control Software that allows the user to define load profile in incremental steps for up to eight control channels. The profile can be paused to allow data collecting, adjust profile, specimen inspection, etc. The profile can be restarted automatically at the end of a user-specified pause or for an indefinite period until user presses the start button. The new software collects and transfers to a graphing application for real-time display and posttest analysis. The graphics software runs on a second computer provided by Gardner Systems. This software is implemented through new hardware provided from Gardner systems. This hardware includes a GS 2100 P program generator with 8 segment generator channels and a bus interface module. There is a GS 2100 D data acquisition unit with a bus interface module. This hardware works together through a GS 2100 PC integration package with a bus interface card, cable, mouse, power control circuit, and a desk top console. To run the software and hardware a new Windows PC with core 2 duo processor, 1 GB memory, 250 GB hard drive, DVD H-RW drive, graphics controller, and printer port will all be provided thru Gardner Systems. This new system, the servo controllers, and signal conditioners are all being serviced on-site. In conjunction with the upgraded system, there are improvements the current researchers would like to see. These include the ability to use Microsoft Excel with the data acquisition software, an emergency power back-up system, a real-time clock, an auto-save function, and a computer based 6 of 12

7 zeroing function for the control boxes. These simple additions to the current upgrade will make it all that easier to use. Conclusions The UET at the University of Houston is receiving an upgrade for the computer software and hardware in the fiscal year. This upgrade makes the use and the versatility of this system much better than it currently is. The availability of user-friendly cyclic loading patterns on element specimens will be available, and will further the research at the Thomas T.C Hsu Structural laboratory. The constitutive laws of more complex concrete designs will be easier to develop with a more flexible system. The upgrade also makes the data acquisition and processing faster and capable to interface with current Microsoft software. This system is the only one of its kind in the world, setting the standard in structural concrete advancement. With the growing need for durable and cost effective infrastructure, the properties of advanced concrete need to be further studied. This will be a reality with the upgraded UET. Acknowledgements The research study described herein was sponsored by the National Science Foundation under the Award No. EEC The opinions expressed in this study are those of the authors and do not necessarily reflect the views of the sponsor. 7 of 12

8 Appendix I. References [1] Hsu, T. T. C., Zhang, L.-X., and Gomez, T., A Servo-Control System for the Universal Panel Tester, Journal of Testing and Evaluations, JTEVA, Vol. 23, No. 6, Nov. 1995, pp [2] Hsu, T. T. C., Belarbi, A., and Pang,X., A Universal Panel Tester, Journal of Testing and Evaluation, JTEVA, Vol. 23, No. 1, January 1995, pp [3] Belarbi, A., T. T. C. Hsu, Constitutive Laws of Reinforced Concrete in Biaxial Tension-Compression, Research Report No. UHCEE 91-2, Department of Civil and Environmental Engineering, University of Houston, Houston, Texas July, 1991, 155 pages. [4] Pang, Xiaobo, T. T. C. Hsu, Constitutive Laws of Reinforced Concrete in Shear, Research Report No. UHCEE 92-1, Department of Civil and Environmental Engineering, University of Houston, Houston, Texas, December, 1992, 180 pages. 8 of 12

9 Appendix II. Figures Figure 1, North view of UET 9 of 12

10 Figure 2, South View of UET 10 of 12

11 Figure 3, Servo Control System 11 of 12

12 Figure 4, LVDT Arrangement 12 of 12