Richard, Attached is the file I used for testing that shows failure as attached.
If I deleted the text in the Introduction section, then tracked changes view okay in the PDF.
At the end of the file I have 3 test lines to see if additional changes still displayed when the Introduction was deleted, and they did display.
I did not incrementally delete parts of the Intro to see when the problem switches on.
The attached exported .tex file does not show lyxadded and lyxdeleted in all caps anywhere.
Hope this provides what you seek. Thanks, Allen ---- Address: Allen Wilkinson (cell) (216) 548-2349 1286 Yellowstone Road Cleveland Heights, OH 44121 USA (INTERNET) aw(at)chaff(dot)biz +++++++ On Wed, 28 No v 2012, Richard Heck wrote:
On 11/27/2012 03:01 PM, Allen Wilkinson wrote:My problem is that if I want to show tracked changes in PDF view or output I get these errors:++++++++++++++++++++++++ Package xcolor Error: Undefined color `LYXDELETED'. Package xcolor Error: Undefined color `LYXADDED'. ... +++++++++++++++++++++++ after a modest finite extent of the document is processed.If I delete changed and unchanged text from the document such that the document is small enough, then PDF output with tracked changes works.This is true whether I use Lyx PDF viewing, or run OS' pdflatex compiler on exported .tex file.If I disable 'show changes in output', then PDF output occurs without errors.Can anyone help me resolve this problem?Odd that the "undefined color" is in all caps. Is LyX exporting this wrong? I think we can only answer this with a test file that causes the problem. rh
cpt.tst.lyx
Description: lyx test file
%% LyX 2.0.5 created this file. For more info, see http://www.lyx.org/. %% Do not edit unless you really know what you are doing. \documentclass[times]{nagauth} \usepackage[T1]{fontenc} \usepackage{xcolor} \usepackage{pdfcolmk} \usepackage{amstext} \PassOptionsToPackage{normalem}{ulem} \usepackage{ulem} \usepackage[unicode=true, bookmarks=true,bookmarksnumbered=true,bookmarksopen=true,bookmarksopenlevel=1, breaklinks=false,pdfborder={0 0 1},backref=false,colorlinks=false] {hyperref} \hypersetup{ colorlinks,bookmarksopen,bookmarksnumbered,citecolor=red,urlcolor=red} \makeatletter %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LyX specific LaTeX commands. \providecolor{lyxadded}{rgb}{0,0,1} \providecolor{lyxdeleted}{rgb}{1,0,0} %% Change tracking with ulem \newcommand{\lyxadded}[3]{{\texorpdfstring{\color{lyxadded}{}}{}#3}} \newcommand{\lyxdeleted}[3]{{\texorpdfstring{\color{lyxdeleted}\sout{#3}}{}}} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% User specified LaTeX commands. % nagdoc.tex V2.0, 13 May 2010 %\documentclass{article} %\documentclass[times,doublespace]{nagauth}%For paper submission \usepackage{moreverb} \usepackage{bm} % TEMP package to exchange messages %\usepackage[dvips,colorlinks,bookmarksopen,bookmarksnumbered,% %citecolor=red,urlcolor=red]{hyperref} \newcommand{\BibTeX}{{\rmfamily B\kern-.05em \textsc{i\kern-.025em b}% \kern-.08em T\kern-.1667em\lower.7ex\hbox{E}\kern-.125emX}} \def\volumeyear{2012} \makeatother \begin{document} \runningheads{Anton Kulchitsky \emph{et~al.}}{\lyxdeleted{Allen Wilkinson}{Sun Nov 25 13:13:12 2012}{Parametrical}\lyxadded{Allen Wilkinson}{Sun Nov 25 13:13:12 2012}{Parametric} study of cone penetration test\dots} \title{Parametrical study of cone penetration test in lunar simulant using discrete element method} \author{Anton~Kulchitsky\affil{1}\corrauth, Allen~Wilkinson\affil{2}, Jerome~B.~Johnson\affil{1} and Paul~Duvoy\affil{1}} %% \address{<\affilnum{1}First author's address %% (in this example it is the same as the third author)\break %% \affilnum{2}Second author's address>} \address{\affilnum{1}University of Alaska Fairbanks, Institute of Northern Engineering, USA\break \affilnum{2}NASA Glenn Research Center, USA} \corraddr{University of Alaska Fairbanks, Institute of Northern Engineering, PO Box 755910, Fairbanks Alaska~99775-5910, USA. Email:~avkulchits...@alaska.edu} %\cgs{<Contract/grant sponsor name (no number)>} %\cgsn{<Contract/grant sponsor name>}{<number>} \cgsn{NASA Lunar Science Institute project ``Scientific and Exploration Potential of the Lunar Poles'' through subcontract 957706 to Johns Hopkins University under NASA contract}{NNA09DB31A} \begin{abstract} \textit{To be written by Anton, Allen, and Jerry at the end} A DEM model parametric study of particle and ensemble physical properties is performed to identify the most important properties for modeling cone penetration and triaxial compression tests. Cone penetration and triaxial compression test data using the silty sand lunar simulant JSC-1a are used in follow-on simulations to select the best property values needed to faithfully reproduce the actual tests. The results of the parametric study and code validation are presented here. \end{abstract} \keywords{discrete element method, cone penetration test, triaxial test, regolith, lunar simulant, JSC-1a} \maketitle %\footnotetext[2]{} \section{Introduction} This paper presents a piece of the efforts needed to address, for space exploration, the problem of a lack of quantitative engineering tools to predict in-situ regolith behavior for excavation, material handling, traction, and foundations to avoid system failures during missions. \textcolor{black}{Cone Penetration Tests (CPT) and triaxial tests are widely used geotechnical characterization tools in the field and lab, respectively. CPT have perhaps the most well developed soil mechanics understanding of any of the soil mechanics field instruments commonly used. Given that} \textcolor{black}{\emph{in-situ}}\textcolor{black}{{} soil states are difficult to reproduce, quantitative field instruments are important. CPT historically has served the trafficability and foundation engineering disciplines, \cite{Shoop93,Mayne} but not the excavation discipline. CPT was the instrument of choice for Apollo as well as Soviet Lunokhod and Luna missions \cite{Costes1973a,Durgunoglu73,Mitchell1974,Cherkasov1973,Leonovich1975,Leonovich71}, motivated by trafficability characterization. The theory and practice have been reviewed numerous times \cite{Schmertmann1978,Perumpral87,Lunne1997a,Mayne2007,Yu1998,Yu06}. Terrestrial practice depends on world-wide soil penetration databases to heuristically correlate tip resistance and sleeve friction to soil parameters used in system designs. A thoughtful micromechanical discussion of soil failure near a penetrating cone can be found in \cite{Ma1994}. This work fully appreciated that soil strength depends in a coupled way on bulk density, fabric (packing structure), and tensorial stress state. Explicitly: cohesion, friction angle, bulk density, fabric, and stress state, bulk shear modulus, surface adhesion, and surface friction angle are not independent parameters, but rather coupled to each other in ways that have no theory that quantitatively and unambiguously describes the coupling. The coupling causes non-uniqueness of solutions (parameter values) to classical soil mechanics soil strength equations. It is this coupling of classical soil mechanical parameters that requires a more fundamental understanding of soil mechanics if modern predictive engineering practice is to be achieved.} The CPT experimental data used here is from a 2.54~$cm$ diameter 60$^{\text{o}}$ full angle cone and friction sleeve of an ``electric friction cone penetrometer'' following an ASTM standard.\cite{ASTMD5778-07} The tip resistance and sleeve shear stress are compared to simulation results. \textcolor{black}{Triaxial tests are laboratory benchmark tests to characterize the full range of possible strengths of a soil as a function of high confining pressures. Here we limit our interest to consolidated undrained triaxial compression tests (CUTC)\cite{ASTMD4767} that are most relevant to dry extraterrestrial silty-sand regoliths that have cohesion. Such tests provide a bound on what is possible for foundations, traction, and excavation that are of interest to space exploration mission planners. Triaxial tests provide, by way of Mohr-Coulomb linear failure theory, quantitative soil parameters, such as friction angle and cohesion that are used in classical soil mechanics equations for system design.\cite{Wilkinson07} The Mohr-Coulomb theory here again does not resolve the coupling of the effective cohesion and friction angles measured.} As an aside, in the bulk material handling area, also important for extraterrestrial exploration, triaxial tests do not provide useful information. Bulk material handling is at low confining pressures and steady shear. Such classical granular flow characterization is best accomplished with ring shear tests.\cite{Jenike19641980reprinting,Rame2009} Discrete Element Method (DEM) modeling of Cone Penetration Tests (CPT) and Triaxial Tests are not new. The DEM CPT work of Ma\cite{Ma1994}, Jiang\cite{Jiang2006}, and Calvetti\cite{Calvetti05} were in two dimensions, while the work of Arroyo and Butlanska\cite{Butlanska09}\cite{Arroyo10} were in three dimensions. Ma and Jiang examined the stress and strain behavior of granular material in the failure zone about the cone, while the work of Arroyo, Calvetti and Butlanska focused on the engineering correlation of classical soil parameters with CPT simulations. Butlanska, in addition, looked at DEM ensemble inhomogeneity of porosity and contact force network. Butlanska and Arroyo confirmed that 3D inhomogeneity and quality of fit to experiment can not be captured by 2D simulations. \{notice triaxial fitting for DEM parameter determination was used by Arroyo\} In the Agricultural Engineering community, Asaf used wedge and plate penetrator and rotating grousered shear plate experiments as the calibration and validation data for DEM simulations.\cite{Asaf07} Again, the work did not investigate the fundamental stress-strain behavior in the soil failure region. All of this past work used the contact physics that follow the work of Cundall and Strack.\cite{Cundall79} Other particle contact models, such as the fundamental solid mechanics-based Hertz-Mindlin model,\cite{Mindlin1953,Mindlin1949,Mindlin1965,Deresiewicz1974a} common in modern codes, have not been used in a study of CPT and Triaxial tests. There is continuing work in the field to add more physical phenomena like several attractive particle contact interactions\textcolor{blue}{\cite{Tomas2009,Walton2008,Johnson-Walton09}} and hysteretic friction contact effects.\textcolor{blue}{\cite{Cole12}} DEM modeling of the triaxial test has been more common than DEM simulations of CPT with diverse motivations. They have included particle property parametric sensitivity studies with ellipsoidal particles\cite{Ng2006a} and more recent sensitivity studies by Uthus\cite{bib:uthus-hopkins-horvli-2008}, crushing effects\cite{Chang2004}, pre-failure\cite{Jardine2001} and failure strain localization into shear bands\cite{Fazekas06}, lunar and Martian simulant strength for excavator and traction response\cite{Alshibli2010,Berry2006,Knuth2012}, particle shape and size effects\cite{Nezami2007b}, alternative boundary conditions\cite{Ng2004}, fabric tensor behavior with principle stress ratio for ellipsoidal particles\cite{Ng2001a}, and fabric response in cyclic triaxial experiments\cite{OSullivan2009}. Some authors have used triaxial tests to establish DEM particle and ensemble parameters to enable simulation of other systems and tests. Arroyo used the technique for terrestrial purposes.\cite{Arroyo10} Others have used the triaxial test-based particle parameter technique with space exploration applications.\cite{Alshibli2010,Berry2006,Knuth2012} \textcolor{black}{To enable interpretation of CPT tests on the Moon and other space bodies we simulate both the CPT and Triaxial tests. We conduct} a particle parameter sensitivity study, along with parameter determination, for the lunar regolith simulant JSC-1a. Parameter sensitivity may be different, depending on the physical experiment being simulated. Both CPT and CUTC tests provide the experimental data. The parameterized DEM is then applied to independent CPT and CUTC data sets to validate the goodness of fits. The material modeled and measured in this work is a simulant of lunar maré volcanic regolith, a well graded silty-sand, called JSC-1a. It \textcolor{black}{was} chosen because of its greater literature base\cite{Alshibli2009c,Arslan2010a,Zeng2010,JSC-1aMSDS05} of geotechnical characterization as compared to, for example, the more recent lunar highlands simulant NU-LHT-2M. JSC-1a possesses a particle size distribution within the wide range measured for returned lunar maré regolith.\cite{Lsb91} Its range of porosity is narrower than the maré regolith. This in part is due to the complex and reentrant particle shapes of maré not present in JSC-1a. The particle mechanical properties, for example Young's Modulus and Poisson's Ratio, are little studied. The only known work is that of Cole and coworkers.\cite{Cole2007a,Cole10,Cole08} No terrestrial simulant captures all the lunar environmental effects on geotechnical behavior. However, for this work terrestrial measurements on a simulant accepted by the lunar community is sufficient for this current proof of capability. The sections of this paper cover the CPT and Triaxial measurements, DEM model description, the CPT and Triaxial simulations for the parametric study\{?\}, CPT and Triaxial simulations to fit experimental data, ...\{To be filled in as sections appear.\} \section{COUPi Discrete Element Model Description} AW \subsection{Overview} Numerical simulations were performed using the COUPi DEM model. The model includes a high performance parallel computational core, graphics, and \lyxadded{Allen Wilkinson}{Sun Nov 25 11:06:19 2012}{interface }scripting capability with the Lua programming language. The software architecture is designed to enable easy modification of the code to suit any material and differing boundary conditions.\lyxadded{Allen Wilkinson}{Tue Nov 27 13:42:23 2012}{\{AW A unique aspect of code is contact mechanics with hysteretic friction, adhesion, complex particle shapes, inter-particle forces at a distance, JKR, etc. We should get these most positive highlights early in the paper.\}} In\lyxadded{Allen Wilkinson}{Tue Nov 27 12:28:15 2012}{ the} discrete element method approach, initially introduced by Cundall\lyxadded{Allen Wilkinson}{Tue Nov 27 12:15:23 2012}{ and} \lyxdeleted{Allen Wilkinson}{Tue Nov 27 12:15:14 2012}{adn} Strack~\cite{Cundall79}, individual bodies and forces between the bodies are model\lyxdeleted{Allen Wilkinson}{Sun Nov 25 13:10:51 2012}{l}ed explicitly. Every body in the COUPi model is a geometrical union of convex rigid shapes, the lowest level primitive indivisible objects.\lyxadded{Allen Wilkinson}{Tue Nov 27 13:42:31 2012}{ \{AW Will ``every body'' case hold when you get optimized? Did we talk of modeling wall structures with simpler tessellations to reduce compute effort? even if coding effort has to increase.\}} These shapes are modeled by dilated points (spheres), segments (capsules), and triangles fo\lyxadded{Allen Wilkinson}{Sun Nov 25 11:15:12 2012}{l}lowing Hopkins~\cite{bib:Hopkins-engcomp-2004}. In the dilation process, an arbitrary shape is dilated by placing the center of a sphere of fixed radius at every point of the surface of the basic shape. This creates a margin around the basic shape. Such margins are selected big enough to ensure that basic shapes never intersect in the model.\lyxadded{Allen Wilkinson}{Tue Nov 27 13:42:38 2012}{ \{AW This word picture is not clear. Either need to add a picture or delete these preceding two sentences, relying on the cited reference to make it clear.\}} This allows the use of more efficient methods for calculation of the intersection between the shapes. These basic shapes also allow the DEM to efficiently model polyhedra~\cite{bib:hopkins-poly-on-chip-2010}. After initialization, COUPi's core calculations are performed by small increment in time, time steps. Every time step the following major computations are performed:\lyxadded{Allen Wilkinson}{Tue Nov 27 13:42:48 2012}{ \{AW Is there anything unique about this scheme? If not, it can be deleted.\}} \begin{enumerate} \item Broad-phase contact detection step. At this stage the set of pairs of convex parts of particles named \textquotedbl{}atoms\textquotedbl{} in the model that may be in contact is created. This set contains only pairs that are suspected to be in a possible contact now or in a near future. \item For every possible contact pair of convex atoms, it is determined if there is a real contact. If the contact exists, the exact geometrical location of this single contact is found as well as its properties such as overlap between bodies or an area of the contact. \item These properties are used to determine the forces on the contacts at this stage as described below in Sec.~\ref{sec:conmech}. For every particle in the system all forces and torques add up forming the forces and torques applied to the particles. All contact forces need to be stored in a hash table and carried over the next time step. \item Forces and torques related\lyxadded{Allen Wilkinson}{Sun Nov 25 12:03:42 2012}{ to} boundary conditions contribute to the forces and torques at this stage. \item Particles velocities and angular velocities are computed using previously calculated forces and torques. \item Velocity related boundary conditions contribute to the velocities and angular velocities. \item New position and orientation of particles are computed based on their velocities. \end{enumerate} The scripting facility is used to create different model\lyxdeleted{Allen Wilkinson}{Sun Nov 25 12:12:39 2012}{led} scenarios such as different regions and physical conditions.\lyxadded{Allen Wilkinson}{Tue Nov 27 13:54:49 2012}{\{AW This preceding sentence seems abstract/vague. Lua scripts define particle shapes and size distributions, material properties, interaction parameters like restitution coefficient, bounding geometries, boundary movements, variables to be logged, ???\}} The motion of the tools is also determined by the scripts.\lyxadded{Allen Wilkinson}{Tue Nov 27 13:54:58 2012}{ \{AW How is the Lua scripting unique from other DEM code interfaces?\}} First non-changed text test \subsection{First added non-change section} \lyxadded{Allen Wilkinson}{Tue Nov 27 14:19:14 2012}{First added changed text test} \end{document}