AI-Enhanced Soil Classification Using Machine Learning Models within the AASHTO Framework

The American Association of State Highway and Transportation Officials (AASHTO) has been accreditig hot mix asphalt testing laboratories under the AASHTO Accreditation Program since 1988. Laboratories must satisfy many quality system criteria, as well as participate in the AASHTO Materials Reference Laboratory (AMRL) on site assessment and proficiency sample programs, in order to receive accreditation from AASHTO. Analysis of AMRL proficiency sample data indicates that accreditation is beneficial to the construction industry and that the AASHTO Accreditation Program (AAP) meets its objective of promoting uniformity in the testing of construction materials.

Bridge foundations must be designed based on acceptable risks of failure. To secure rapid implementation of Load Resistance Factor Design (LRFD) principles for foundation design, the American Association of State Highway and Transportation Officials (AASHTO) and the Federal Highway Administration (FHWA) are requiring their use through AASHTO code. The Bridge Section of the Oregon Department of Transportation (ODOT) has responsibility for satisfactory design of all the bridge structures across the state’s highway system. The widespread geotechnical adoption of the LRFD code throughout state DOTs has been difficult in the case of deep foundations due to regional differences and in some cases a lack of any close match to DOT foundation practices. This lack of matching stems from the source research conducted on which the code is based, documented as NCHRP 507. For ODOT, the evaluation of nominal axial static capacity for each driven pile in the field is conducted by dynamic methods and AASHTO offers resistance factors for these techniques. ODOT typically uses the wave equation software (WEAP) applied at the end of initial driving, EOID, and occasionally at the beginning of pile restrike (BOR) to capture increases in capacity from set-up. This study reports that, based on past and new surveys, ODOT practice is reasonably typical for DOT practice in sands, silts, and clays. The AASHTO resistance factor, φ, for WEAP is at EOID and is too low for the efficient design of piles to match the likely probabilities of pile failure. The survey of Northwest state DOTs revealed that 80% of the DOTs believe that a φ of 0.4 is conservative and 37.5 % do not use the AASHTO-sanctioned φ of 0.4. Matching LRFD to allowable stress design (ASD) by direct calibration for a single pile, without any reported capacity bias, sets φ as 0.55 to match the ASD factor of safety of 2.5. An ODOT case history of a recently completed pilesupported bridge designed and constructed to FHWA and AASHTO ASD standards in use at that time, shows the number of piles at the bent studied would be doubled under new AASHTO requirements. This suggests the standard will add considerable pile foundation costs to all new bridges. This cost increase is a strong incentive to complete statistical recalibration of GRLWEAP dynamic capacity resistance value in a Phase 2 of this study.

This paper evaluates the American Association of State Highway and Transportation Officials (AASHTO) line load and live load distribution factors. The standard AASHTO formulas for live load in some cases underestimate the live load moments by as much as 40%. The AASHTO distribution of line loads (dividing the loads equally among all girders) results in even larger underestimates of line load distribution, leading to unsafe designs. This paper shows how more reliable distribution factors can be obtained for line loads, AASHTO truck loads, and non-AASHTO live loads. A user-friendly computer program for calculating distribution factors is discussed. A useful concept for analyzing beams of variable section (equivalent girder) is also discussed.

Current pipeline design practice within water, wastewater, and drainage pipe industries is largely based on the American Water Works Association (AWWA) design manuals and standards, American Society for Testing and Materials (ASTM) standards and American Association of State Highway and Transportation Officials (AASHTO) codes and design practices, respectively. AWWA, ASTM and AASHTO practices recognize the importance of proper specification and installation of bedding and backfill for achieving the intended structural performance of buried rigid and more importantly flexible pipelines. However, the guidance provided by various AWWA design manuals, specific for pipe materials, ASTM standards, and AASHTO codes use different terminology and requirements to classify and specify bedding and backfill materials based on the pipe material. This variability of bedding and backfill classification and specification requirements creates unnecessary confusion amongst pipeline designers and introduces a potential for unintended use of bedding and backfill materials. More importantly, the use of inconsistent and non-standard soil classification terminologies makes the interdisciplinary communications difficult, potentially resulting in inappropriate and inconsistent selection of engineering properties for the pipeline design. With an understanding of the potential risks from these different specifications based on pipe material, the development of a uniform classification of bedding and backfill materials has been proposed in the past. Due care must be exercised in selecting such a uniform classification system for bedding and backfill. This paper presents the opinion that such a uniform classification and specification system should be based upon fundamental soil mechanics principles, the Unified Soil Classification System (USCS), and engineering properties of soils. The system should also consider the past and present practices of pipeline design. In an attempt to develop such a system, this paper presents the following: 1) A summary of the current practices of classification and specification of bedding and backfill for water/wastewater and storm drainage pipelines. 2) A comparison of current classification and specification practices of bedding and backfill for different pipe materials noting common and differentiating parameters. 3) A proposed uniform framework for classification and specification based on USCS and engineering properties of the bedding and backfill materials.

Evaluation of subgrade soil stiffness provides valuable information about the expected behavior and response of pavements, which can be used to group pavement sections with similar structural capacities and pavement maintenance and rehabilitation to prioritize projects. A deflection test, such as the falling weight deflectometer (FWD), is commonly used to determine subgrade resilient modulus (MR) from its measured deflection data. A correction factor has been applied to the American Association of State Highway and Transportation Officials (AASHTO) MR equation for adjusting the difference between laboratory MR and the in situ measurements (with the FWD). However, large variations on the correction factor result in less accuracy of subgrade MR based on FWD measurements. The objective of this research is to evaluate the accuracy of the AASHTO MR equation and improve its accuracy by adjusting the MR correction factor. This paper consists of subtasks: (a) evaluating the current AASHTO MR correction factor; (b) evaluating the affecting factors (i.e., subgrade soil stiffness, bedrock depth, highway classification, and environmental condition); and (c) developing a framework to determine accurate MR correction factor.

Construction of any sustainable road requires the determination of engineering properties of the materials to be used in the project. In this work, the impact of the construction materials used in some parts on Anambra State (southeastern Nigeria) with significant road pavement failure rate is investigated. Soil and aggregate samples were collected in situ on some failed spots on major roads in the state and were subjected to standard laboratory analyses. The results indicated that the roads were constructed on sandy subgrades with gravel-sized particles that ranged from 0 – 43 %, sand 25 – 77 %, and fines 14 – 63 %. The liquid limit of the subgrade materials ranged from 27 – 44 %, plasticity index 10 – 24 %, while the linear shrinkage ranges from 8 – 11 %. These Atterberg limit results denote low to moderate expandable soils. Maximum dry density ranged between 1.93 - 2.20 Mg/m3, while optimum moisture content ranged from 6.78 to 15.9%. The California bearing ratio (CBR), on the other hand, ranged between 60 – 161 %, for unsoaked condition, and 17.8 %, for soaked condition, indicating fairly stable materials. Based on the American Association of State Highway and Transportation Officials (AASHTO) classification system, the subgrade samples classify mainly as A-2-6 soil or as CL according to the Unified Soil Classification System (USCS). The particle density value ranges from 2.35 - 2.71 Mg/m3, while the field relative compaction density ranges between 89 – 108 %. Analysis on the sub base and base course aggregates indicated bulk density of fine aggregate (sand) that ranged from 0.16 - 1.68 Mg/m3. Comparison of the results with an engineering standard revealed that the materials did not meet standard limits(FMW&H), hence, the failure recorded on the paved roads. Water inflow would result in subgrade material expansion, hence, stability and bearing capacity reduction. ANOVA test result indicates significant variable from the data source while principal component and correlation analyses suggests that moisture content, liquid limit and plasticity index have direct statistical relationships. This research therefore advocates proper drainage design and stabilization of the subgrade materials in the area during road construction.

Optimization of prestressed concrete bridge girder section using a modified harmony search algorithm

It is well established that vehicular traffic traveling over bridge-like structures can impart a dynamic load effect that is greater than vehicles’ static weight alone. In order to account for this increased load, bridge design codes use a factor known as the dynamic load allowance (IM) to amplify static vehicular live loads. In the current version of the American Association of State Highway and Transportation Officials (AASHTO) Manual for Bridge Evaluation (MBE), reductions in IM are allowed for bridges having span lengths greater than 12.2 m with road surfaces in good condition. In addition, the current AASHTO LRFD Bridge Design Specifications allow for a reduced IM for culverts with higher fill depth. However, many culverts have neither span lengths greater than 12.2 m nor higher fill depths and thus are not eligible for such IM reductions. This paper investigated whether similar IM reductions can be considered for culverts with smaller span lengths and fill depths. The field experiments conducted suggest that culverts having span lengths less than 12.2 m and fill depths less than 0.5 m could be considered for similar IM reductions allowed by the MBE.

The Applied Technology Council (ATC), in a joint venture with the Multidisciplin ary Center for Earthquake Engineering Research (MCEER), has recently developed Comprehensive Specifications for the Seismic Design of Highway Bridges (National Cooperative Highway Research Program, NCHRP Project 12-49). The American Association of State Highway and Transportation Officials (AASHTO) is considering these recommended specifications for possible incorporation into the future AASHTO Load and Resistance Factor Design Specifications. The primary objective of NCHRP Project 12-49 was to develop seismic design provisions that reflect the latest research findings, design philosophies, and design approaches (ATC 2002). In this study, the substructures of two existing Missouri bridges with Seismic Performance Category of “C” and “D” were redesigned according to the proposed NCHRP provisions. These bridges have different resisting systems; in one bridge abutments contribute in lateral resisting system and in the second bridge, only the intermediate bent resist against earthquake forces. The design res ults were compared for these bridges.

Calcium-based stabilizers (CaO-based stabilizers) develop a high pH environment when mixed with water during construction. This high pH condition and the presence of lime (CaO) introduce an environment favorable to the formation of two expansive minerals, ettringite and possibly thaumasite, when sulfate-bearing salts are present in soil. The success of a lime stabilization project depends heavily on the accuracy of predictions regarding the potential of the native soils to form these disruptive minerals. This report describes the nature of these minerals and the mechanism of their formation in soils stabilized with lime, cement, or fly ash. The key issue for the engineer in deciding how to properly stabilize sulfate-bearing soils is to quickly and efficiently determine: (1) the sulfate content of the soils and (2) the threshold quantity of sulfates likely to cause damage. This report addresses both issues but focuses on the methods used to quantify sulfate content. Quick and effective field techniques and more precise laboratory tests are compared and evaluated. Recommendations are presented regarding the appropriate testing protocol to use. Specific recommendations for changes to the American Association of State Highway and Transportation Officials (AASHTO) test method T-290-95 (methods A and B) are made. The report also outlines techniques for effective risk assessment including reconnaissance, topographical assessment, pedological and geological assessment, climatic assessment, and soil sampling plans. The report also outlines construction tests that can reduce risks for various levels of sulfate determined using a method such as AASHTO T-290-95 (revised) or equivalent.

Impact in highway prestressed concrete bridges

A pavement mechanistic-empirical analysis is based on a pre-designed structure checked for required performance criteria. In case the latter are not met, this structure is modified and reprocessed. In this context, analyzing the effect of variations in project parameters on pavement performance prediction subsidizes a better understanding of results provided by computer programs. The objective of this study is to assess the effect of layer thickness and resilience modulus variations on flexible pavement performance. To do so, performance was estimated for the 20th project year through Elastic Layered System Model 5 (ELSYM5) software and American Association of State Highway and Transportation Officials (AASHTO) Mechanistic-Empirical method (ME). Using multiple regression models for result adjustment and through statistical assessments on regression coefficients calculated, it can be concluded that pavement lifespan consumption, predicted by simulations on ELSYM5, is sensitive to variations in coating and subbase thickness and in subgrade resilience modulus. For AASHTO ME method, predicted values for distresses were significantly sensitive to variations in coating, base and subbase thickness, and in base and subgrade resilience modulus. Comparing both approaches, it is concluded that ELSYM5 can be a viable alternative to the application of a ME pavement design method.

The Texas State Department of Highways and Public Transportation uses the Interactive Graphics roadway design system (IGrds), an AASHTOWARE product of the American Association of State Highway and Transportation Officials (AASHTO). IGrds provides roadway designers two optional methods of operation: one with an extensive set of computation and graphic capabilities through a graphic work station; and the other through input of the data using an alphanumeric terminal and editor and performing the calculations in a noninteractive (background) mode. The department is enhancing IGrds to provide automated plan preparation capabilities to minimize the work of their engineers and draftspersons to produce construction plans. The AASHTO IGrds running on stand‐alone work stations is an integral part of the effort. The Automated Plan Preparation System will be the link that brings together the design systems and minimizes manual drafting of construction plans. The ability to transfer “intelligent” graphics from a desi...

Variable Message Sign (VMS) systems are widely used in motorways to provide traffic information to motorists. Such systems are subjected to wind-induced structural vibration that can lead to damage due to fatigue. The limited information that is available on the safe wind design of VMS motivated a large scale testing that was conducted at the Wall of Wind (WOW) Experimental Facility at Florida International University (FIU). One of the objectives of the present study was to experimentally assess the wind-induced force coefficients on VMS of different geometries and utilize these results to provide improved design guidelines. A comprehensive range of VMS geometries were tested and mean normal and lateral force coefficients, in addition to the twisting moment coefficient and eccentricity ratio, were determined using the measured data for each model, for wind directions of 0o and 45o. The results confirmed that the mean drag coefficient on a prismatic VMS is smaller than the value of 1.7 suggested by American Association of State Highway and Transportation Officials (AASHTO). An alternative to this value is presented in the form of a design matrix with coefficients ranging from 0.98 to 1.28, depending on the aspect and depth ratio of the VMS. Furthermore, results indicated that the corner modification on a VMS with chamfered edges demonstrated a reduction in the drag coefficient compared to sharper edges. Finally, the dynamic loading effects were considered by evaluating the gust effect factor, using the ASCE 7 formulations, for various VMS weights and geometries. The findings revealed a wide range of possible gust effect factors, both above and below the current AASHTO specification of 1.14. Future research may include different geometries of VMS and a wider range of wind directions.

In designing and assessing pavement performance, the uncertainty in material properties and geometrical variables of pavement and in traffic and environmental actions should be considered. A single factor is employed to deal with these uncertainties in the current American Association of State Highway and Transportation Officials (AASHTO) guide for design of pavements. However, use of this single factor may not ensure reliability-consistent pavement design and assessment because different random variables that may have different degrees of uncertainty affect the safety and performance of pavement differently. Similar problems associated with structural design have been recognized by code writers and dealt with using partial safety factors or load resistance factors. The present study is focused on evaluating a set of partial safety factors to be used in conjunction with the flexible pavement deterioration model in the Ontario pavement analysis of cost and the model in the AASHTO guide for evaluating the flexible pavement performance or serviceability. Evaluation and probabilistic analyses are carried out using the first-order reliability method and simple simulation technique. The results of the analysis were used to suggest factors that could be used, in a partial safety factor format, for designing or assessing flexible pavement conditions to achieve a specified target safety level.Key words: deterioration, reliability, pavement, serviceability, stochastic process, performance, partial safety factor.