Main Lemma Repository
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

lemma.h 14KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336
  1. /* This file is part of Lemma, a geophysical modelling and inversion API */
  2. /* This Source Code Form is subject to the terms of the Mozilla Public
  3. * License, v. 2.0. If a copy of the MPL was not distributed with this
  4. * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
  5. /**
  6. @file
  7. @author Trevor Irons
  8. @date 06/19/2009 09:12:20 AM The Birth of Lemma!
  9. @version $Id: lemma.h 203 2015-01-09 21:19:04Z tirons $
  10. **/
  11. // \image html lemma.png
  12. /** \mainpage Lemma is an ElectroMagnetics Modelling API
  13. \authors Trevor Irons and M. Andrew Kass and others
  14. Originally Lemma was intended as a recursive acronym standing for
  15. <B>L</B>emma is an <B>E</B>lectro<B>M</B>agnetics <B>M</B>odelling <B>A</B>PI.
  16. As the breadth of the project has expanded, the name has remained
  17. appropriate in a more literal sense. Lemma is a flexible cross-platform
  18. library delivering an expressive API that can be used to easily create
  19. versatile programs. Lemma is not itself a program, instead it is a
  20. collection of building blocks to make geophysical applications.
  21. We retain this name because:
  22. - In mathematics a Lemma is a proven proposition which is used as a
  23. stepping stone to a larger result rather than as a statement in-and-of
  24. itself.
  25. - In addition to the electromagnetic modelling, some other facilities are
  26. provided such as numerical optimization and inversion capabilities. These
  27. tools are also considered stepping stones to final products.
  28. We feel that this is a particularly approprate name, as Lemma's
  29. API can be leveraged create powerful applications such as forward
  30. modelling and inverting frequency and time-domain
  31. surveys of arbitrary survey design, sNMR surveys, CSAMT and more.
  32. \section Motivation
  33. Why another Geophysical EM project? For starters, there aren't that many
  34. quality open source packages out there. Those that do exist are generally
  35. specialized to perform a single task and extending them is a major undertaking.
  36. Lemma's approach is much different, by providing a set of general tools users
  37. can easily assemble applications that suite their needs. Furthermore, most are
  38. written in either Fortran or MATLAB, and can be difficult to integrate into
  39. multiphysics applications. To our knowlege, Lemma is the only C++ EM simulation
  40. package freely available.
  41. \section Capabilities Capabilities
  42. In the long term, we have many goals for this software project. Due to its
  43. design, Lemma can be built upon and extended easily. The initial aim is to
  44. provide flexible 1D and 3D EM modelling in the time and frequency domains.
  45. The project is still in beta, but we have made a lot of progress already.
  46. We will release our first non-beta release as soon as the following are
  47. supported.
  48. \subsection FDM Frequency-domain forward modelling
  49. Lemma was initially called EMMODFD: Electromagnetic Modelling in the Frequency
  50. Domain. As such this is the most mature area of Lemma.
  51. \par 1D
  52. Frequency domain solutions to electrical and magnetic dipoles can be computed
  53. quasi-analytically in 1D. Calculations can be made in or above the layered
  54. media, and complex electrical conductivity and magnetic susceptibility are
  55. supported according to the Cole-Cole model. Sources may be embdedded in the
  56. media or in the resisitive air layer. Lemma can also can compute fields due
  57. to arbitrarily shaped ungrounded wire loops, topography of the loops is also
  58. supported. Two separate approaches to solving the Hankel transform, one
  59. based on Anderson's digitial filtering technique, and another based on Gaussian
  60. quadrature.
  61. \par 3D
  62. A fast 3D solver that can modify the 1D results based
  63. on arbitrary electrical conductivity model is nearing completion and is
  64. provided in a separate module.
  65. \par future work
  66. We are also planning on supporting grounded wires in the near future.
  67. \subsection TDM Time-domain forward modelling
  68. A 1D time-domain solution has been implemented that utilises both a
  69. dipole source as well as a wire loop. Currently, only one receiver is
  70. modelled at a time, but will be generalised. In addition, utilities
  71. to read in data files for modelling have been implemented.
  72. We would like to offer 3D time domain support, but this will not be
  73. provided before our first stable release.
  74. \subsection DataFormats Data Formats
  75. The EM community is plagued with myriad data formats. Often each equiptment
  76. manufacturer provides their own data format and interoperability is a
  77. constant struggle. We are working on a flexible data format based on the XML
  78. format that can be adapted to many types of data. The template for this
  79. format will be publically released and we hope it catches on in the community.
  80. At the least, it will provide a mechanism to compare datasets and datatypes
  81. within Lemma.
  82. \section Modules Modules
  83. Due to Lemma's design, it is easy to extend the platform. In some cases this
  84. extension results in adding functionality that is not directly related to
  85. ElectroMagnetics. The following modules utilise parts of Lemma to provide
  86. their functionality.
  87. \section Tutorials
  88. - \ref Tutorial - Basic intruduction to Lemma, including aquiring and
  89. compiling the code, class structure, and building your own
  90. applications.
  91. - \ref Extending Tutorial on how to extend Lemma.
  92. \section Development Development and design
  93. Ths package was initially developed by the Center for Gravity, Electrical, and
  94. Magnetic Studies (CGEM) at the Colorado School of Mines (CSM), the United
  95. States Geological Survey (USGS), and Broken Spoke Development, LLC. Where it drew
  96. on work by many others including Ki Ha Lee, and Walt Anderson. All new work and
  97. interfaces are written entirely in C++. Several small external projects are
  98. included, which are written in standard C, and FORTRAN 77. We adapt a
  99. modern, test driven, object oriented, C++ framework.
  100. More recent development has been undertaken at the University of Utah through the Energy
  101. and Geoscience Institute.
  102. \section Legalities
  103. \subsection Copyrights
  104. The following copyrights apply to the source.
  105. Most of the code was developed either by Trevor
  106. Copyright (C) 2008-2010 Trevor Irons <trevor.irons@lemmasoftware.org> or
  107. M. Andrew Kass Copyright (C) 2010 <mkass@usgs.gov>.
  108. The 1D EM solver was derived (but updated heavily) from a Fortran
  109. programme written by Ki Ha Lee in 1984. We have communicated with Ki Ha,
  110. and he assured us that this code is in the public domain.
  111. A Gaussian quadrature hankel transform originally written by Alan Chave was
  112. ported to C++. This code is in the public domain, and the source code was
  113. published in Geophysics.
  114. A digital filtering approach to the Hankel transform written by Walt
  115. Anderson was also rewritten for Lemma. The original Fortran code is also in
  116. the public domain.
  117. Please note that Ki Ha Lee and Walt Anderson had no part in this work, and
  118. the above should not be interpreted as any sort of endorsement by those
  119. parties.
  120. \subsection License
  121. This Source Code Form is subject to the terms of the Mozilla Public
  122. License, v. 2.0. If a copy of the MPL was not distributed with this
  123. file, You can obtain one at http://mozilla.org/MPL/2.0/.
  124. \section Contributing Suggestions and contributions
  125. We welcome contributions and suggestions. Feel free to email the development
  126. team at info@lemmasoftware.org.
  127. Under the terms of the MPL, if you modify a Lemma file, you are obligated to
  128. share those contributions back with the community.
  129. \section Useful Useful links
  130. - Home page https://lemmasoftware.org
  131. - Git repository https://git.lemmasoftware.org
  132. - Broken Spoke Develpment http://numericalgeo.com
  133. - CGEM at the Coloroado School of Mines http://geophysics.mines.edu/cgem/
  134. - EGI at the Eniversity of Utah http://egi.utah.edu/
  135. **/
  136. #pragma once
  137. #ifndef __LEMMA_H
  138. #define __LEMMA_H
  139. #include <LemmaConfig.h>
  140. // Include some basic stuff that will always be needed
  141. #include <iostream>
  142. #include <iomanip>
  143. #include <complex>
  144. #include <fstream>
  145. #include <string>
  146. #include <vector>
  147. #include <stdexcept>
  148. #include <sstream>
  149. #include <Eigen/Core>
  150. #include <cstddef>
  151. #include <Eigen/StdVector>
  152. #include <Eigen/Sparse>
  153. #include <unsupported/Eigen/FFT>
  154. //#include <unsupported/Eigen/SparseExtra>
  155. #include <Eigen/Geometry>
  156. /** \brief The only namespace used by Lemma
  157. *
  158. * \details The rational behind this namespace is that built-in
  159. * types should be used wherever possible, but not
  160. * not built-in names. This allows for code that is better
  161. * enacsulated and easier to modify. The typedefs and constants
  162. * specified here are defined so that
  163. * precision/inplimentation can easily be changed.
  164. * All floating precision types should be typedefed in this file
  165. * and should not be used natively within any code.
  166. * Lemma uses
  167. * the Eigen Matrix/Vector/Linear Algebra library.
  168. * <http://eigen.tuxfamily.org> and a lot of the namespece typedefs
  169. * are specifying Eigen types.
  170. */
  171. namespace Lemma {
  172. /// Real defines precision for the whole API, default is double
  173. #ifdef LEMMA_SINGLE_PRECISION
  174. typedef float Real;
  175. #else // ----- LEMMA_SINGLE_PRECISION -----
  176. typedef double Real;
  177. #endif // ----- not LEMMA_SINGLE_PRECISION -----
  178. /// Complex version of Real.
  179. typedef std::complex<Real> Complex;
  180. /// A 3 component Eigen vector of Reals
  181. typedef Eigen::Matrix<Real, 3, 1> Vector3r;
  182. /// A 3 X Dynamic Component Eigen matrix of Reals
  183. typedef Eigen::Matrix<Real, 3, Eigen::Dynamic> Vector3Xr;
  184. /// Variable length Eigen vector of Reals
  185. typedef Eigen::Matrix<Real, Eigen::Dynamic, 1> VectorXr;
  186. /// Variable length Eigen vector of integers (int)
  187. typedef Eigen::Matrix<int, Eigen::Dynamic, 1> VectorXi;
  188. /// Variable length Eigen vector of Complexes
  189. typedef Eigen::Matrix<Complex, Eigen::Dynamic, 1> VectorXcr;
  190. /// A 3 Component Eigen vector of Complexes
  191. typedef Eigen::Matrix<Complex, 3, 1> Vector3cr;
  192. /// A 3 X Dynamic Component Eigen matrix of Complexes
  193. typedef Eigen::Matrix<Complex, 3, Eigen::Dynamic> Vector3Xcr;
  194. /// Variable length Eigen Matrix of Reals
  195. typedef Eigen::Matrix<Real, Eigen::Dynamic, Eigen::Dynamic> MatrixXr;
  196. /// Variable length Eigen Matrix of ints
  197. typedef Eigen::Matrix<int, Eigen::Dynamic, Eigen::Dynamic> MatrixXi;
  198. /// Variable length Eigen vector of Complexes
  199. typedef Eigen::Matrix<Complex, Eigen::Dynamic, Eigen::Dynamic> MatrixXcr;
  200. ////////////////////////////////////////
  201. // Constants used across the programmes
  202. /// Restating the obvious, this is pi
  203. const Real PI = 4.0*atan(1.0);
  204. /// Permitivity of Free Space
  205. //const Real EPSILON0 = 8.854187817e-12;
  206. const Real EPSILON0 = 8.854187817e-12;
  207. /// Permeability of free space
  208. const Real MU0 = 4.*PI*1e-7;
  209. /// 1/4 of \f$ \pi\f$
  210. const Real QPI = .25/PI;
  211. /// Some functions will convert units from SI (standard) to Gauss
  212. /// This is because NMR calculations are much more natural in Gauss
  213. enum MAGUNITS {TESLA, NANOTESLA, GAUSS};
  214. /// Unit of temperature entered
  215. enum TEMPUNITS {CELCIUS, KELVIN};
  216. /// Unit of time entered
  217. enum TIMEUNITS {SEC, MILLISEC, MICROSEC, NANOSEC, PICOSEC};
  218. /// Unit of time entered
  219. enum FREQUENCYUNITS {HZ, KHZ, MHZ, GHZ};
  220. /// FEM coil relative orientations
  221. enum FEMCOILORIENTATION {COAXIAL, COPLANAR};
  222. /// General orientation relative to coordinate system
  223. enum ORIENTATION {X, Y, Z, NX, NY, NZ};
  224. /// Type of field
  225. enum FIELDTYPE {HFIELDREAL, HFIELDIMAG, EFIELDREAL, EFIELDIMAG};
  226. /// Compenent of vector field
  227. enum FIELDCOMPONENT {XCOMPONENT=0, YCOMPONENT=1, ZCOMPONENT=2};
  228. /// Spatial component of vector
  229. enum SPATIALCOORDINANT {XCOORD=0, YCOORD=1, ZCOORD=2};
  230. /** Evaluation method for Hankel integrals.
  231. * ANDERSON801 Walt Anderson's 801 point filter
  232. * CHAVE Alan Chave's gaussian quadrature integration method
  233. * FHTKEY201 Key's 201 point filter
  234. * FHTKEY201 Key's 101 point filter
  235. * FHTKEY51 Key's 51 point filter
  236. * QWEKEY Key's Gaussian quadrature integration method
  237. */
  238. enum HANKELTRANSFORMTYPE { ANDERSON801, CHAVE, FHTKEY201, FHTKEY101, FHTKEY51, QWEKEY,
  239. FHTKONG61, FHTKONG121, FHTKONG241, IRONS };
  240. /** Enum is OK because these are the only physically possible sources.
  241. @param NOSOURCETYPE is default.
  242. @param ELECTRICDIPOLE is an electric dipole
  243. @param MAGNETICDIPOLE is a magnetic dipole
  244. */
  245. enum DipoleSourceType {NOSOURCETYPE, GROUNDEDELECTRICDIPOLE, UNGROUNDEDELECTRICDIPOLE, MAGNETICDIPOLE};
  246. /// Only three polarizations are supported. They may be summed to
  247. /// approximate others
  248. /// @param NOPOLARISATION is uninitialized, default value
  249. /// @param XPOLARISATION is a dipole oriented in the x direction
  250. /// @param YPOLARISATION is a dipole oriented in the y direction
  251. /// @param ZPOLARISATION is a dipole oriented in the z direction
  252. enum DipoleSourcePolarisation {NOPOLARISATION, XPOLARISATION,
  253. YPOLARISATION, ZPOLARISATION};
  254. /// The polarity may be either negative or positinve
  255. enum DipoleSourcePolarity {NEGATIVE, POSITIVE};
  256. /** The fields to make calculations on
  257. */
  258. enum FIELDCALCULATIONS {E, H, BOTH};
  259. /** Windowing function type
  260. */
  261. enum WINDOWTYPE { HAMMING, /*!< A hamming window */
  262. HANNING, /*!< A hanning window */
  263. RECTANGULAR /*!< Rectangular window */
  264. };
  265. }
  266. #endif // __Lemma_H