Gongkang Fu

High-accuracy edge detection with Blurred Edge Model

A high-accuracy edge detection algorithm at sub-pixel level is proposed in this work. A blurred edge model is adopted here, and a least-squared-error based solution is derived. Its applications to both synthetic and real images are presented for evaluation. It is compared with two other sub-pixel edge detectors. One uses a moment-based approach, and the other an interpolation-based approach. The comparison shows higher accuracy of the proposed algorithm, even for image data with significant noise. An application of the proposed algorithm in engineering is also introduced herein.

Truck Loads and Bridge Capacity Evaluation in China

Over the past 2–3 decades, the economic development in China has natured the establishment of a highway network with a large number of bridges. However, there are still no nationwide specification provisions for assessing their safe load carrying capacity. The research work reported herein focuses on developing reliability based requirements for this purpose. In this study, weigh-in-motion data for more than 7.3 million trucks were gathered from highways in three provinces of China, continuously over 1–16 months in 2006 and 2007. The data were processed and projected to model the live-load spectrum over 3-year and 100-year periods, respectively. The former is the required bridge inspection interval and the latter the bridge design life span, according to current Chinese maintenance and design specifications. The proposed projection method is shown to be more reliable compared with those reported. The resulting load spectra are used to assess the structural reliability of typical Chinese highway bridges at the component level. Based on the accordingly selected target reliability index, the live-load factors for bridge evaluation are developed in this study, proposed to be included in the Chinese national specifications.

Edge-Based Close-Range Digital Photogrammetry for Structural Deformation Measurement

This paper proposes a close-range digital photogrammetric system based on edge detection for structural deformation measurement. Different from traditional photogrammetric applications using discrete points, continuous edges in digital images are used here as the controlling feature in this new system. This makes it possible to acquire spatially intensive information. The system uses several digital images of the structure, taken from a number of different stations before and after the deformation is induced. Then, an image-matching algorithm based on the coplanarity condition developed here is applied to establish spatial relationships of the interested edges, identified by using a new high-precision method. These relationships are subsequently employed to acquire full-field deformation measurement of the structure. A series of experiments was conducted in the laboratory to investigate the capability of the new photogrammetric system. Results show that the system is highly accurate and suitable for structural deformation measurement. It offers noncontact, full-field, and spatially intensive measurement, in contrast with conventional contact and point-measurement.

Extrapolation for Future Maximum Load Statistics

The use of short-term load data to extrapolate or project remote future maximum load has been practiced in structural design code development and/or calibration. However, this approach has not been adequately evaluated or validated because of the absence of long-term data, which is important for the structures’ safety governed by these codes. In this paper, the application of such extrapolations to highway bridge design is studied by taking advantage of available long-term truck weight data obtained from the weigh-in-motion technique. A new extrapolation method is proposed from the understanding of the importance of the load probability distribution’s high tail. This approach significantly reduces the mathematical length of the extrapolation/projection and thus increases its reliability. Also presented in this paper are application examples. In addition, the extrapolation process proposed is evaluated with quantitative indexes. The application of a temporal extrapolation/projection can now be evaluated by...

Inspection and monitoring techniques for bridges and civil structures

Testing steel corrosion in reinforced concrete Alkali-silica reaction (ASR) testing of corrosion in concrete Acoustic testing of concrete Electrical impedance testing of wood components Detecting decay in wood components Testing timber pile length in bridges Ultrasonic testing of structural timber components Digital radioscopy analysis of timber structures Visual inspection techniques for bridges and other transport structures Acoustic emission testing of bridges Bridge inspection using virtual reality and photogrammetry Discontinuity in masonry wall.

Multiple Presence Factor for Truck Load on Highway Bridges

Trucks may simultaneously appear on a highway bridge in the same lane or different lanes, which represent the governing load for short- and medium-span bridges. In practice specifications, the multiple presence factor (MPF) to cover these simultaneous loads was developed using in- tuitionormodelswithoutsupportofmeasuredweightdataoftrucksinmotion.ThispaperpresentsaresearcheffortofderivingaMPFbasedonweigh- in-motion(WIM)truck data, forboththestrengthandfatiguelimitstates,includingabout 68milliontrucksgathered for436monthsandfrom 43 sites in California, Oregon, Michigan, and New York. The resulting MPFs are proposed in this paper as functions of bridge-span length, truck-traffic vol- ume, and number of lanes available. They show that the code-specified MPF values are conservative and sometimes overconservative by 400% or more, which may have caused too high requirements for load-rating existing bridges, especially for shorter spans and low truck traffic. The recom- mended MPF may be considered to be adopted in specifications for highway bridge design and evaluation. This paper also presents a new truck-by- truckanalysisapproachforunderstandingtheloadeffectof trucks in motion,takingadvantageof increasinglyavailableWIMdataof largequantityto avoidunsupportedassumptionsinlive-loadmodeling.DOI:10.1061/(ASCE)BE.1943-5592.0000330.©2013 American Society of Civil Engineers. CE Database subject headings: Trucks; Maximum loads; Fatigue; Highway bridges; Design. Author keywords: Truck load; Maximum load; Fatigue load; Highway bridge; Design and evaluation; Multiple presence.

Elastically supported cantilever beam subjected to nonstationary seismic excitation

The dynamic behaviour of bridge piers under seismic load is studied here in the context of random vibration. The earthquake excitation is modelled as white noise filtered by the Clough–Penzien filter in cascade with modulation accounting for intensity non-stationarity. The bridge pier modelled as an elastically supported cantilever beam witha lumped mass at the top. An analytical solution is presented for the response statistics, which may be used to develop probabilistic seismic response spectra for design. It is found that the first two modes of the pier approach to rigid-body motion when the stiffness of the elastic support decreases. Seismic responses increase with the top mass, resultingin significantly high displacement and shear but negligible moment at the top, and higher shear and moment at thebase. Lower stiffness of the elastic support increases the pier top displacement and moment responses, but may increase or reduce shear responses. The probabilistic spectrum of the relative displacement between the bridge superstructure and the pier top may depend on the two systems’ relative modal properties.

Planning for Commercial Vehicle Weight Limit Change: Application and Computer Software

Commercial vehicle weight limits here refer to those on the gross weight, axle spacing, and axle weights. They have a notable influence on the deterioration and lifespan of highway facilities such as pavements and bridges. A new methodology is presented here for estimating the monetary impact to a network of highway bridges due to commercial vehicle weight limit increase. This methodology focuses on four categories of cost impact: (1) life reduction due to fatigue of steel components; (2) life reduction due to fatigue of reinforced concrete decks; (3) requirements for addressing deficiency of existing bridges; and (4) capacity requirements for new bridges. For illustration, an application example of the new methodology is presented. The results show that Categories 3 and 4, additional inadequate bridges and increased requirements for new bridges, contribute more significantly to the total impact cost. In addition, a computer software program for the methodology has been developed and is introduced in this paper to facilitate implementation. It is intended to be used for highway planning at different network levels, such as for state, provincial, federal, or national jurisdictions.

LADAR for structural damage detection

LADAR here stands for laser radar, using laser reflectivity for measurement. This paper presents a new technique using LADAR for structure evaluation. It is experimentally investigated in the laboratory. With cooperation of the US Federal Highway Administration, a recently developed LADAR system was used to measure structural deformation. The data were then treated for reducing noise and used to derive multiple features for diagnosis. The results indicate a promising direction of nondestructive evaluation using LADAR.

Bridge Safety Monitoring Using: Coupled Current Device Images

A new method of global diagnosis for bridge safety monitoring using high-resolution coupled current device (CCD) images is presented. Its advantages are as follows: (a) no sensors need to be attached to the structure, (b) no mathematical modeling is required (such as finite element methods), (c) a probability-based diagnosis approach is used to deal with measurement noise associated with image data, and (d) CCD cameras are affordable now and are expected to continue to drop in price, which will make this new technique workable for implementation. The new method was applied in the laboratory to a bridge structure model, as presented. Experimental results show that the method is able to identify stiffness losses as small as 3 percent for both their presence and their location. It is concluded that the proposed approach is promising for experimentation in the field.