Browsing by Author "Scott, Michael L."

Now showing 1 - 3 of 3
  • Thumbnail Image
    Automated Characterization of Bridge Deck Distress Using Pattern Recognition Analysis of Ground Penetrating Radar Data
    Scott, Michael L. (Virginia Tech, 1999-08-02)
    Many problems are involved with inspecting and evaluating the condition of bridges in the United States. Concrete bridge deck inspection and evaluation presents one of the largest problems. The deterioration of these concrete decks progresses more rapidly than any other bridge component, which leads to early concrete deck replacements that must be done before the bridge superstructure needs to be replaced. The primary cause of deterioration in these concrete bridge decks is corrosion-induced concrete cracking, which frequently results in delaminations. Delamination distress increases the life cycle cost of maintaining a concrete bridge deck, particularly when it is not detected early on. Early detection of delamination distress can facilitate economical repair and rehabilitation work, but bridge engineers must recommend deck replacement if repairs are delayed too long or inspection tools cannot detect delaminations early enough. The Federal Highway Administration has responded to the need for a better bridge deck inspection tool by contracting Lawrence Livermore National Laboratory to develop two new prototype ground penetrating radar systems. These two systems generate three-dimensional data that provide a representation of features that lie below the bridge deck surface. Both of these systems produce large amounts of data for an individual bridge deck, which makes automated data processing very desirable. The primary goal of the automated processing is to characterize bridge deck distress represented in the data. This study presents data collected from sample bridge deck sections using one of the prototype systems. It also describes the development and implementation of appropriate methods for automating data processing. The automated data processing is accomplished using image processing and pattern recognition algorithms developed in the study.
  • Thumbnail Image
    Compiler-Directed Failure Atomicity for Nonvolatile Memory
    Liu, Qingrui; Izraelevitz, Joseph; Lee, Se Kwon; Scott, Michael L.; Noh, Sam H.; Jung, Changhee (Department of Computer Science, Virginia Polytechnic Institute & State University, 2019-07-15)
    This paper presents iDO, a compiler-directed approach to failure atomicity with nonvolatile memory. Unlike most prior work, which instruments each store of persistent data for redo or undo logging, the iDO compiler identifies idempotent instruction sequences, whose re-execution is guaranteed to be side effect-free, thereby eliminating the need to log every persistent store. Using an extension of prior work on JUSTDO logging, the compiler then arranges, during recovery from failure, to back up each thread to the beginning of the current idempotent region and re-execute to the end of the current failure-atomic section. This extension transforms JUSTDO logging from a technique of value only on hypothetical future machines with nonvolatile caches into a technique that also significantly outperforms state-of-the art lock-based persistence mechanisms on current hardware during normal execution, while preserving very fast recovery times.
  • Thumbnail Image
    Determining the Air Void Parameters of Concrete Using Digital Image Analysis of Polarized Light Micrographs
    Scott, Michael L. (Virginia Tech, 1997-04-22)
    The ASTM C457 test has long been a standard used to obtain the air void parameters of concrete materials. These air void parameters provide valuable information that has been linked to the performance of concrete under conditions such as freezing and thawing cycles. The standard test procedure involves linearly traversing a cut and polished section of a concrete specimen while a technician observes it under a microscope. Chord lengths of material constituents that the technician observes along the linear traverse are recorded and later used to calculate air void parameters statistically. This procedure is long and tedious, which makes it susceptible to human error due to operator fatigue. This study proposes and implements a new test method for evaluating concrete air void parameters using an image analysis method. A polishing procedure along with a differential interference contrast microscope are used to obtain high contrast images of material constituents, which provide raw data for the image analysis method. Because of the high contrast that can be obtained, cement paste, air voids in the cement paste, and aggregate materials in the concrete can be distinguished from one another based on these images. An image analysis program has been written for this study which linearly traverses these images and records the chord lengths of material constituents in a similar way to the standard ASTM C457 test. The chord length data must be processed further, however, because features in the images can be truncated by the edge of the image. Correction calculations for this problem are implemented in the image analysis algorithm. Two specimens which have been previously tested using the standard ASTM C457 method by the Virginia Transportation Research Council, (VTRC), are used in this study. The air void parameters obtained using the new test are compared directly with the results obtained by VTRC for the two specimens. Statistical comparisons indicate that the results of the new test are indeed significant, showing the potential it has for practical implementation. There are drawbacks to the test including a long polishing procedure, but this process can be automated. The new test appears to have excellent potential for practical application, but it should be emphasized that the test has only been implemented using materials in two concrete specimens. Further study on a variety of other concrete materials would be required for implementation in a standard procedure.