Cliff ord Algebras Daniel Klawitter Cliff ord Algebras Geometric Modelling and Chain Geometries with Application in Kinematics Foreword by Prof. Dr. Gunter Weiss Dr. Daniel Klawitter Dresden, Germany Dissertation TU Dresden, 2014 ISBN 978-3-658-07617-7 ISBN 978-3-658-07618-4 (eBook) DOI 10.1007/978-3-658-07618-4 Th e Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografi e; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de. Library of Congress Control Number: 2014953940 Springer Spektrum © Springer Fachmedien Wiesbaden 2015 Th is work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifi cally the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfi lms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, compu- ter soft ware, or by similar or dissimilar methodology now known or hereaft er developed. Exempted from this legal reservation are brief excerpts in connection with reviews or schol- arly analysis or material supplied specifi cally for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law. Th e use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specifi c statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal re- sponsibility for any errors or omissions that may be made. Th e publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Springer Spektrum is a brand of Springer DE. Springer DE is part of Springer Science+Business Media. www.springer-Spektrum.de To my parents, Manuela and Burghard Foreword What is the right mathematical model to a real phenomenon of our world? Do there exist criteria whether a model can be called eleg- ant as well as practically efficient? These questions will surely be answered differently by e.g. a pure mathematician on the one hand and an engineer on the other. But both will have to start with ab- stracting real world phenomena to objects of a more or less platonic world. Inthisidealizedworldoneconstructsthe”structuredgeomet- ricimage”ofrealworldprocessesandobjectsinconsideration. Thus, on the way to constructing explicit mathematical models, geometric representations,andgeometricreasoningwillalwaysplayanessential partandprovidesinsightintoperhapsnotthatobviousinterrelations and structures. The idea for this book took its origins from space kinematics and robotics and the standardized way. Its aim is to be understood and handledpractically. Therearetwoessentiallydifferentwaystomodel space kinematics, although both ways are based on an Euclidean mo- tion as the core element of the manifold of motions. Loosely spoken, a seemingly clear model will interpret such a basic element as a point and the manifold of motions as a point manifold. (It turns out, that, via a clever transfer mapping, this point mani- fold becomes a hyper-quadric in a seven-dimensional real projective space. This hyper-quadric is referred to as Study’s quadric.) The second model interprets a motion as a number rather than a point on a line: In order to describe the position of it on that line, one needs a pair of Hamilton quaternions as coordinates, which in addition fulfill conditions. Both models relate to well understood families of models. Thefirstonecanbeseene.g. asahigherdimensionalanaloguetothe geometry of lines and screws. Indeed, screws are closely connected to motions. The second model relates to so-called circle geometries. VIII Foreword Their structures are those of a line with coordinates of a ring. (A standard example of such a circle geometry is the so-called (planar) Mo¨bius geometry; here it turns out that the coordinate ring of the corresponding line is even a field, namely the field of complex num- bers.) Both standard models for space kinematics have their merits and can be considered elegant. The first model aims at a visualization of the structured set of motions, where coordinates become inessential. In spite of the transfer between the original set of motions and Study’s quadric, the seven-dimensional real projective space is described by coordinates. Now the model induces objects of further research: For example, planar intersections of Study’s quadric lead to interesting subsets of motions. The second model aims at a clearly perform- able calculus and the analytic treatment of forced motions. Here as well, one feels the need for visualization. Such transfer from calculus to geometry establishes a third model and can be derived as follows: Pairsofquaternionsareinterpretedaselementsofafour-dimensional module over the ring of so-called dual numbers, which again are in- terpreted as points. Points originating from motions are contained by a hypersphere in this four-dimensional module space and, again, planar intersections of this sphere become a matter of interest. But these planar intersections stem from helical motions in the original world. In this setting, space kinematics and its models are used to point out an obvious fact: The models and model spaces, which are struc- turedfortheirown,influencethetopicsofresearchoftheoriginalreal world objects and ” re-structure ” these objects in a specific manner. Therefore, the answer to the introductory question, which model is best, will be rather a matter of taste and pre-knowledge of related models than being based on objectified facts. But still the origin of different models is the object of consideration! So the question about relations between different models and their structures arises. Is is possible to unify these view points? This last question is the core topic of this book: The author in- troduces the seemingly universal tool of Clifford Algebras and their Foreword IX groupstructurestomodelnotonlyEuclideankinematicsandlinegeo- metry, but also circle geometries and their generalizations, so-called ” chain geometries ” and, to take it even further, also non-Euclidean kinematics and line geometry. Projective and chain geometric prop- erties alternately relate to algebraic properties and concepts. Again, the concept ” cross ratio ” becomes such an interface between Geo- metry and Algebra, being on the one hand visualized as quadruplets of points and on the other being a number resp. a ring element as well as responsible for structuring e.g. specific point sets of a line to chains. The presented method allows to connect disciplines of (Applied) Mathematics, that, up to now, are treated independently. Furthermore, it provides an effective tool for solving practical prob- lems of Mechanical Engineering and Robotics, too. Is this book a Maths book on theoretical and practical applications of Clifford Algebras? Yes, but in its rich details and examples it aims at visualizing the more or less hidden Geometries of the Clifford Al- gebra calculus, and therefore, it could rightly be called a Geometry book. By its wide range of examples it is profitable to a broad span of readers, be it mechanical engineers or mathematicians. Like me, they surely consider its publication as an absolute stroke of luck. Wien Gunter Weiss Preface Usually, dual unit quaternions are used to describe Euclidean dis- placements in three-dimensional kinematics. Engineers widely use this elegant calculus next to a real projective geometric model called Study’s quadric. Moreover, there exists a ” sphere-model ” in a four- dimensional module space over the ring of dual numbers. However, there are no investigations connecting these three models up to now. After recalling Study’s quadric, the ” sphere-model ”, and other mod- els,wepresentasurvey-likeintroductiontoCliffordalgebrasandtheir Spin and Pin groups. The advantage of Clifford algebras is that geo- metric objects and transformation may be represented as algebra ele- ments. Using the so-called sandwich operator, it is possible to apply transformations that are represented by algebra elements directly to geometric objects, also described by algebra elements. We introduce the homogeneous and conformal Clifford algebra model. Further- more, we show how to model special Cayley-Klein geometries and their isometry groups in a homogeneous Clifford algebra model. As anexample, wefocusonthehomogeneousCliffordalgebramodelcor- responding to line geometry and derive the correspondence between projective transformations of three-dimensional projective space and the Pin group Pin . Therefore, we introduce the Clifford algebra (3,3,0) C(cid:2) , constructedoverthequadraticspaceR(3,3), anddescribehow (3,3,0) points on Klein’s quadric are embedded as null vectors. We discuss how geometric entities that are known from Klein’s model, i.e., lin- ear line manifolds can be transferred to this homogeneous Clifford algebra model. All entities known from line geometry occur natur- ally in this model and can be transformed projectively by the ap- plication of the sandwich operator. The action of grade-1 elements correspondstothe actionof null polarities onP3(R), i.e., correlations that are involutions as the basic elements building up the group of regular projective transformations. It is proven that every regular projective transformation of P3(R) can be expressed as the product XII Preface of six null polarities, i.e., skew-symmetric 4×4 matrices at the most. The results achieved for Klein’s quadric may be transferred to any quadric, hence, we present the homogeneous Clifford algebra model corresponding to Lie sphere geometry as an example. Additionally, a new geometric algebra allowing the description of inversions with re- specttoquadricsinprincipalpositionasPingroupispresented. This model serves as a generalization for the conformal geometric algebra and is constructed for dimension two and three in detail. Further- more, the generalization to arbitrary dimension is shown. A further focus of the thesis is applying chain geometry to Clifford algebras in order to examine the cross ratio in Clifford algebras. It is well known that the cross ratio of four complex numbers is real if, and only if, they all lie on a Mo¨bius circle, i.e., a circle or a line augmented by a point at infinity. A generalization of the so called M¨obius geometry is obtained by using different algebras instead of complex numbers. This leads to a branch of geometry called chain geometry. Chains are subsets of the projective line over a ring which can be parametrised with the cross ratio. Therefore, it is natural to apply this theory to dual quaternions and to examine the kinematic and geometric interpretation. A more general point of view can be achieved by the use of Clifford algebras and Spin groups instead of dual quaternions and dual unit quaternions. After recalling the fun- damental chain geometric background, we define the cross ratio for CliffordalgebrasandtheirPinandSpingroups. Wepresentaquadric model corresponding to the dual unit quaternions and homogeneous Clifford algebra models of Klein’s, Study’s and Lie’s quadric where chainsthatarecontainedinthegrade-1subspacecorrespondtoconic sections. Moreover,wederiveanalgebraicbiarcconstructionwiththe help of contact spaces. Chains of the grade-1 subspace that are in contact at a certain point are parametrized with the cross ratio and correspond to conic sections. Moreover, it has been proven that the connected components of the Pin- and Spin groups define subspaces of chain spaces. Every element contained in a chain defined by three elements of the same connected component of the Pin- or the Spin group is contained in the same connected component of the Pin- or the Spin group. The question for the cross ratio of dual unit qua- ternions has been answered in detail.
Description: