
<p>Proceedings of the Institute of Mathematics and Mechanics, National Academy of Sciences of Azerbaijan Volume 44, Number 2, 2018, Pages 328–337 </p><p>DIAGONAL LIFTS OF METRICS TO COFRAME BUNDLE </p><p>HABIL FATTAYEV AND ARIF SALIMOV </p><p>Abstract. In this paper the diagonal lift <sup style="top: -0.3013em;">D</sup>g of a Riemannian metric g of a manifold M<sub style="top: 0.1245em;">n </sub>to the coframe bundle F<sup style="top: -0.3013em;">∗</sup>(M<sub style="top: 0.1245em;">n</sub>) is defined, Levi-Civita connection, Killing vector fields with respect to the metric <sup style="top: -0.3012em;">D</sup>g and also an almost paracomplex structures in the coframe bundle are studied. </p><p>1. Introduction </p><p>The Riemannian metrics in the tangent bundle firstly has been investigated by the Sasaki [14]. Tondeur [16] and Sato [15] have constructed Riemannian metrics on the cotangent bundle, the construction being the analogue of the metric Sasaki for the tangent bundle. Mok [7] has defined so-called the diagonal lift of metric to the linear frame bundle, which is a Riemannian metric resembles the Sasaki metric of tangent bundle. Some properties and applications for the Riemannian metrics of the tangent, cotangent, linear frame and tensor bundles are given in [1-4,7-9,12,13]. This paper is devoted to the investigation of Riemannian metrics in the coframe bundle. In 2 we briefly describe the definitions and results that are needed later, after which the diagonal lift <sup style="top: -0.3298em;">D</sup>g of a Riemannian metric g is constructed in 3. The Levi-Civita connection of the metric <sup style="top: -0.3299em;">D</sup>g is determined in In 4. In 5 we consider Killing vector fields in coframe bundle with respect to Riemannian metric <sup style="top: -0.3299em;">D</sup>g. An almost paracomplex structures in the coframe bundle equipped with metric <sup style="top: -0.3299em;">D</sup>g are studied in 6. </p><p>2. Preliminaries </p><p>We shall summarize briefly the basic definitions and results which be used later. <br>Let M<sub style="top: 0.1363em;">n </sub>be an n−dimensional differentiable manifold of class C<sup style="top: -0.3299em;">∞ </sup>and F<sup style="top: -0.3299em;">∗</sup>(M<sub style="top: 0.1363em;">n</sub>) its coframe bundle (see, [10, 11]). The coframe bundle F<sup style="top: -0.3299em;">∗</sup>(M<sub style="top: 0.1363em;">n</sub>) over M<sub style="top: 0.1363em;">n </sub>consists of all pairs (x, u<sup style="top: -0.3298em;">∗</sup>), where x is a point of M<sub style="top: 0.1364em;">n </sub>and u<sup style="top: -0.3298em;">∗ </sup>is a basis (coframe) for the cotangent space T<sup style="top: -0.3299em;">∗</sup>M. We denote by π the natural projection of F<sup style="top: -0.3299em;">∗</sup>(M<sub style="top: 0.1363em;">n</sub>) to M<sub style="top: 0.1363em;">n </sub>defined by π(x, u<sup style="top: -0.3299em;">∗</sup><sup style="top: -0.8545em;">x</sup>) = x. If (U; x<sup style="top: -0.3299em;">1</sup>, x<sup style="top: -0.3299em;">2</sup>, ..., x<sup style="top: -0.3299em;">n</sup>) is a system of local coordinates in M<sub style="top: 0.1363em;">n</sub>, then a coframe u<sup style="top: -0.3299em;">∗ </sup>= (X<sup style="top: -0.3299em;">α</sup>) = (X<sup style="top: -0.3299em;">1</sup>, X<sup style="top: -0.3299em;">2</sup>, ..., X<sup style="top: -0.3299em;">n</sup>) for T<sub style="top: 0.2248em;">x</sub><sup style="top: -0.3299em;">∗</sup>M<sub style="top: 0.1363em;">n </sub>can be expressed uniquely in the form X<sup style="top: -0.3299em;">α </sup>= X<sub style="top: 0.2547em;">i</sub><sup style="top: -0.3299em;">α</sup>(dx<sup style="top: -0.3299em;">i</sup>)<sub style="top: 0.1364em;">x </sub>and hence </p><p></p><ul style="display: flex;"><li style="flex:1">ꢀ</li><li style="flex:1">ꢁ</li></ul><p></p><p></p><ul style="display: flex;"><li style="flex:1">π</li><li style="flex:1"><sup style="top: -0.3753em;">−1</sup>(U); x<sup style="top: -0.3753em;">1</sup>, x<sup style="top: -0.3753em;">2</sup>, ..., x<sup style="top: -0.3753em;">n</sup>, X<sub style="top: 0.2248em;">1</sub><sup style="top: -0.3753em;">1</sup>, X<sub style="top: 0.2248em;">2</sub><sup style="top: -0.3753em;">1</sup>, ..., X<sub style="top: 0.2248em;">n</sub><sup style="top: -0.3753em;">n </sup></li></ul><p></p><p>2010 Mathematics Subject Classification. 53C25, 55R10. </p><p>Key words and phrases. Riemannian metric, coframe bundle, diagonal lift, Levi-Civita connection, Killing vector field, almost paracomplex structure. </p><p>328 </p><ul style="display: flex;"><li style="flex:1">DIAGONAL LIFTS OF METRICS TO COFRAME BUNDLE </li><li style="flex:1">329 </li></ul><p></p><p>is a system of local coordinates in F<sup style="top: -0.3299em;">∗</sup>(M<sub style="top: 0.1363em;">n</sub>) (see, [10]). Indices i, j, k, ..., α, β, γ, ... have range in {1, 2, ..., n}, while indices A, B, C, ... have range in </p><p>ꢂ</p><p></p><ul style="display: flex;"><li style="flex:1">1, ..., n, n + 1, ..., n + n<sup style="top: -0.3753em;">2 </sup></li><li style="flex:1">.</li></ul><p></p><p>We put h<sub style="top: 0.1363em;">α </sub>= α · n + h. Summation over repeated indices is always implied. </p><p>r</p><p>We denote by =<sub style="top: 0.2248em;">s</sub>(M<sub style="top: 0.1364em;">n</sub>) the set of all differentiable tensor fields of type (r, s) on M<sub style="top: 0.1363em;">n</sub>. Let V = V <sup style="top: -0.3299em;">i</sup>∂<sub style="top: 0.1363em;">i </sub>and ω = ω<sub style="top: 0.1363em;">i</sub>dx<sup style="top: -0.3299em;">i </sup>be the local expressions in U ⊂ M<sub style="top: 0.1363em;">n </sub>of a </p><p></p><ul style="display: flex;"><li style="flex:1">1</li><li style="flex:1">0</li></ul><p></p><p>vector and a covector (1-form) fields V ∈ =<sub style="top: 0.2436em;">0</sub>(M<sub style="top: 0.1364em;">n</sub>) and ω ∈ =<sub style="top: 0.2436em;">1</sub>(M<sub style="top: 0.1364em;">n</sub>), respectively. Then the complete and horizontal lifts <sup style="top: -0.3298em;">C</sup>V,<sup style="top: -0.3298em;">H </sup>V ∈ =<sub style="top: 0.2436em;">0</sub>(F<sup style="top: -0.3298em;">∗</sup>M<sub style="top: 0.1364em;">n</sub>) of V and the β−th </p><p>1</p><p>V<sub style="top: 0.1172em;">β </sub></p><p>1</p><p>vertical lifts ω ∈ =<sub style="top: 0.2436em;">0</sub>(F<sup style="top: -0.3299em;">∗</sup>M<sub style="top: 0.1363em;">n</sub>) (β = 1, 2, ..., n) of ω have respectively, </p><p></p><ul style="display: flex;"><li style="flex:1">ꢃ</li><li style="flex:1">ꢄ</li><li style="flex:1">ꢃ</li><li style="flex:1">ꢄ</li><li style="flex:1">ꢃ</li><li style="flex:1">ꢄ</li></ul><p></p><p>V <sup style="top: -0.3299em;">i </sup><br>V <sup style="top: -0.3299em;">i </sup></p><p>0</p><p>V<sub style="top: 0.1172em;">β </sub></p><p></p><ul style="display: flex;"><li style="flex:1"><sup style="top: -0.3754em;">C</sup>V = </li><li style="flex:1">,</li><li style="flex:1"><sup style="top: -0.3754em;">H</sup>V = </li><li style="flex:1">,</li><li style="flex:1">ω = </li></ul><p></p><p>(2.1) </p><p>X<sub style="top: 0.2547em;">j</sub><sup style="top: -0.3298em;">α</sup>Γ<sup style="top: -0.4423em;">j</sup><sub style="top: 0.2936em;">ik</sub>V <sup style="top: -0.3298em;">k </sup></p><ul style="display: flex;"><li style="flex:1">δ<sub style="top: 0.2768em;">β</sub><sup style="top: -0.3298em;">α</sup>ω<sub style="top: 0.1364em;">i </sub></li><li style="flex:1">−X<sub style="top: 0.2547em;">j</sub><sup style="top: -0.3299em;">α</sup>∂<sub style="top: 0.1364em;">i</sub>V <sup style="top: -0.3299em;">j </sup></li></ul><p></p><p></p><ul style="display: flex;"><li style="flex:1">n</li><li style="flex:1">o</li></ul><p></p><p></p><ul style="display: flex;"><li style="flex:1">∂</li><li style="flex:1">∂</li></ul><p></p><p>with respect to the natural frame {∂<sub style="top: 0.1363em;">i</sub>, ∂<sub style="top: 0.1363em;">i </sub>} = details). </p><p>,</p><p>, (see [10] for more </p><p>∂X<sub style="top: 0.2208em;">i</sub><sup style="top: -0.2288em;">α </sup></p><p>∂x<sup style="top: -0.1892em;">i </sup></p><p>α</p><p>The vertical lift of a smooth function f on M<sub style="top: 0.1364em;">n </sub>is a function <sup style="top: -0.3298em;">V </sup>f on F<sup style="top: -0.3298em;">∗</sup>(M<sub style="top: 0.1364em;">n</sub>) defined by <sup style="top: -0.3299em;">V </sup>f = f ◦ π. <br>Let (U, x<sup style="top: -0.3298em;">i</sup>) be a coordinate system in M<sub style="top: 0.1364em;">n</sub>. In U ∈ M<sub style="top: 0.1364em;">n</sub>, we put </p><p>∂<br>∂x<sup style="top: -0.2627em;">i </sup></p><p>X<sub style="top: 0.185em;">(i) </sub></p><p>=</p><p></p><ul style="display: flex;"><li style="flex:1">,</li><li style="flex:1">θ<sup style="top: -0.3753em;">(i) </sup>= dx<sup style="top: -0.3753em;">i</sup>, i = 1, 2, ..., n. </li></ul><p></p><p>H</p><p>V<sub style="top: 0.083em;">α </sub>(i) </p><p>Taking account of (2.1), we easily see that the components of X<sub style="top: 0.185em;">(i) </sub>and </p><p>θ</p><p>are respectively, given by </p><p></p><ul style="display: flex;"><li style="flex:1">ꢃ</li><li style="flex:1">ꢄ</li></ul><p>ꢄ</p><p>δ<sub style="top: 0.2547em;">i</sub><sup style="top: -0.3298em;">h </sup></p><p></p><ul style="display: flex;"><li style="flex:1">ꢀ</li><li style="flex:1">ꢁ</li></ul><p></p><p>D<sub style="top: 0.1364em;">i </sub>= <sup style="top: -0.3753em;">H</sup>X<sub style="top: 0.185em;">(i) </sub>= A<sup style="top: -0.3753em;">H</sup><sub style="top: 0.2247em;">i </sub></p><p>=</p><p>,</p><p>(2.2) (2.3) </p><p>X<sub style="top: 0.2547em;">j</sub><sup style="top: -0.3299em;">α</sup>Γ<sup style="top: -0.4423em;">j</sup><sub style="top: 0.2936em;">ih </sub></p><p>ꢃ</p><ul style="display: flex;"><li style="flex:1">ꢀ</li><li style="flex:1">ꢁ</li></ul><p></p><p>0</p><p>δ<sub style="top: 0.2767em;">β</sub><sup style="top: -0.3299em;">α</sup>δ<sub style="top: 0.2767em;">h</sub><sup style="top: -0.3299em;">i </sup></p><p>V<sub style="top: 0.083em;">α </sub>(i) </p><p>D<sub style="top: 0.1364em;">i </sub></p><p>=</p><p>θ</p><p>= A<sup style="top: -0.3753em;">H</sup><sub style="top: 0.2247em;">i </sub></p><p>=</p><p>αα</p><p>ꢂ</p><p>with respect to the natural frame {∂<sub style="top: 0.1364em;">i</sub>, ∂<sub style="top: 0.1364em;">i </sub>}. We call the set <sup style="top: -0.3298em;">H</sup>X<sub style="top: 0.185em;">(i)</sub>, <sup style="top: -0.3298em;">V</sup><sup style="top: -0.2468em;">α </sup>θ<sup style="top: -0.3298em;">(i) </sup>the </p><p>α</p><p>frame adapted to the Levi-Civita connection ∇<sub style="top: 0.1363em;">g</sub>. On putting </p><p>V<sub style="top: 0.083em;">α </sub>(i) </p><p>D<sub style="top: 0.1364em;">i </sub>= <sup style="top: -0.3753em;">H</sup>X<sub style="top: 0.185em;">(i)</sub>, D<sub style="top: 0.1364em;">i </sub></p><p>=</p><p>θ , </p><p>α</p><p>we write the adapted frame as {D<sub style="top: 0.1408em;">I</sub>} = {D<sub style="top: 0.1363em;">i</sub>, D<sub style="top: 0.1363em;">i </sub>}. From equations (2.2), (2.3) </p><p>α</p><p>V<sub style="top: 0.083em;">α </sub></p><p>and (2.1) we see that <sup style="top: -0.3299em;">H</sup>V and ω have respectively, components </p><p></p><ul style="display: flex;"><li style="flex:1">ꢃ</li><li style="flex:1">ꢄ</li></ul><p></p><p>0</p><p>δ<sub style="top: 0.1183em;">α</sub><sup style="top: -0.4423em;">β</sup>ω<sub style="top: 0.1364em;">i </sub></p><p>V <sup style="top: -0.3299em;">i </sup></p><p>0</p><p>ꢃ</p><ul style="display: flex;"><li style="flex:1">ꢀ</li><li style="flex:1">ꢁ</li></ul><p></p><p><sup style="top: -0.3754em;">H</sup>V = V <sup style="top: -0.3754em;">i</sup>D<sub style="top: 0.1363em;">i</sub>, <sup style="top: -0.3754em;">H</sup>V = <sup style="top: -0.3754em;">H</sup>V <sup style="top: -0.3754em;">I </sup></p><p>=</p><p>,</p><p>(2.4) (2.5) </p><p>ꢄ<br>X</p><ul style="display: flex;"><li style="flex:1">ꢀ</li><li style="flex:1">ꢁ</li></ul><p></p><p></p><ul style="display: flex;"><li style="flex:1">ω = </li><li style="flex:1">ω<sub style="top: 0.1363em;">i</sub>δ<sub style="top: 0.2248em;">α</sub><sup style="top: -0.3754em;">β</sup>D<sub style="top: 0.1363em;">i </sub>, ω = </li></ul><p></p><p>V<sub style="top: 0.0831em;">α </sub></p><p>ω<sup style="top: -0.3754em;">I </sup></p><p>=</p><p></p><ul style="display: flex;"><li style="flex:1">V<sub style="top: 0.0831em;">α </sub></li><li style="flex:1">V<sub style="top: 0.0831em;">α </sub></li></ul><p></p><p>α</p><p>i</p><p>with respect to the adapted frame {D<sub style="top: 0.1407em;">I</sub>}, where V <sup style="top: -0.3298em;">i </sup>and ω<sub style="top: 0.1364em;">i </sub>being local components </p><p></p><ul style="display: flex;"><li style="flex:1">1</li><li style="flex:1">0</li></ul><p></p><p>of V ∈ =<sub style="top: 0.2436em;">0</sub>(M<sub style="top: 0.1363em;">n</sub>) and ω ∈ =<sub style="top: 0.2436em;">1</sub>(M<sub style="top: 0.1363em;">n</sub>), respectively. <br>Let us consider local 1−forms η˜<sup style="top: -0.3299em;">I </sup>in π<sup style="top: -0.3299em;">−1</sup>(U) defined by </p><p>η˜ = A <sub style="top: 0.2248em;">J </sub>dx<sup style="top: -0.3753em;">J </sup>, </p><p></p><ul style="display: flex;"><li style="flex:1">I</li><li style="flex:1">I</li></ul><p></p><p>¯where </p><p></p><ul style="display: flex;"><li style="flex:1"> </li><li style="flex:1">!</li><li style="flex:1"> </li><li style="flex:1">!</li></ul><p></p><p></p><ul style="display: flex;"><li style="flex:1">i</li><li style="flex:1">i</li></ul><p></p><p>δ<sub style="top: 0.2547em;">j</sub><sup style="top: -0.3299em;">i </sup></p><p>0</p><ul style="display: flex;"><li style="flex:1">¯</li><li style="flex:1">¯</li></ul><p></p><p>A <sub style="top: 0.2547em;">j </sub>A <sub style="top: 0.2547em;">j </sub></p><p>I</p><p>β</p><p></p><ul style="display: flex;"><li style="flex:1">A<sup style="top: -0.3753em;">−1 </sup>= (A <sub style="top: 0.2247em;">J </sub>) = </li><li style="flex:1">=</li></ul><p></p><p>.</p><p>(2.6) <br>¯</p><p>−X<sub style="top: 0.2248em;">m</sub><sup style="top: -0.3299em;">α </sup>Γ<sup style="top: -0.3299em;">m</sup><sub style="top: 0.2546em;">ij </sub>δ<sub style="top: 0.2767em;">β</sub><sup style="top: -0.3299em;">α</sup>δ<sub style="top: 0.2714em;">i</sub><sup style="top: -0.4423em;">j </sup></p><p>i<sub style="top: 0.083em;">α </sub></p><p>j</p><p>i<sub style="top: 0.083em;">α </sub></p><p>¯</p><p>A</p><p>¯</p><p>A</p><p>j<sub style="top: 0.1172em;">β </sub></p><p></p><ul style="display: flex;"><li style="flex:1">330 </li><li style="flex:1">HABIL FATTAYEV AND ARIF SALIMOV </li></ul><p></p><p>The matrix (2.6) is the inverse of the matrix </p><p></p><ul style="display: flex;"><li style="flex:1"> </li><li style="flex:1">!</li><li style="flex:1"> </li><li style="flex:1">!</li></ul><p></p><p></p><ul style="display: flex;"><li style="flex:1">j</li><li style="flex:1">j</li></ul><p></p><p>γ</p><p>j<sub style="top: 0.1172em;">β </sub></p><p>δ<sub style="top: 0.2936em;">k</sub><sup style="top: -0.4423em;">j </sup><br>X<sub style="top: 0.1183em;">m</sub><sup style="top: -0.4423em;">β </sup>Γ<sup style="top: -0.3299em;">m</sup><sub style="top: 0.2768em;">jk </sub>δ<sub style="top: 0.1183em;">γ</sub><sup style="top: -0.4423em;">β</sup>δ<sub style="top: 0.2546em;">j</sub><sup style="top: -0.3299em;">k </sup></p><p>0</p><p></p><ul style="display: flex;"><li style="flex:1">A<sub style="top: 0.2936em;">k </sub></li><li style="flex:1">A<sub style="top: 0.2936em;">k </sub></li></ul><p></p><p></p><ul style="display: flex;"><li style="flex:1">A = (A<sub style="top: 0.2247em;">K </sub><sup style="top: -0.3754em;">J </sup>) = </li><li style="flex:1">=</li><li style="flex:1">(2.7) </li></ul><p></p><p>j<sub style="top: 0.1172em;">β </sub></p><p></p><ul style="display: flex;"><li style="flex:1">A<sub style="top: 0.2935em;">k </sub></li><li style="flex:1">A<sub style="top: 0.2935em;">k </sub></li></ul><p></p><p>γ</p><p>of the transformation D<sub style="top: 0.1407em;">K </sub>= A<sub style="top: 0.2694em;">K </sub><sup style="top: -0.3299em;">J </sup>∂<sub style="top: 0.1407em;">J </sub>( see (2.2) and (2.3)). It is easy to establish </p><p>ꢂ</p><p>that the set η˜<sup style="top: -0.3299em;">I </sup>is the coframe dual to the adapted frame {D<sub style="top: 0.1408em;">K</sub>}, i.e. η˜ (D<sub style="top: 0.1408em;">K</sub>) = A <sub style="top: 0.2248em;">J </sub>A<sub style="top: 0.2247em;">K </sub>= δ<sub style="top: 0.2247em;">K</sub><sup style="top: -0.3753em;">I </sup>. </p><p></p><ul style="display: flex;"><li style="flex:1">I</li><li style="flex:1">I</li><li style="flex:1">J</li></ul><p></p><p>¯</p><p>D</p><p>3. Diagonal lift g of a Riemannian metric g to the coframe bundle </p><p>On putting locally </p><p>n</p><p>X</p><p></p><ul style="display: flex;"><li style="flex:1"><sup style="top: -0.3754em;">D</sup>g = g<sub style="top: 0.1363em;">ij</sub>η˜<sup style="top: -0.3754em;">i </sup>⊗ η˜<sup style="top: -0.3754em;">j </sup>+ δ<sub style="top: 0.1481em;">αβ </sub></li><li style="flex:1">g<sup style="top: -0.3754em;">ij</sup>η˜<sup style="top: -0.3754em;">i</sup><sup style="top: 0.0831em;">α </sup>⊗ η˜<sup style="top: -0.3754em;">j</sup><sup style="top: 0.1172em;">β </sup></li></ul><p></p><p>(3.1) </p><p>i,j </p><p>in the coframe bundle F<sup style="top: -0.3299em;">∗</sup>(M<sub style="top: 0.1363em;">n</sub>), we see that <sup style="top: -0.3299em;">D</sup>g defines a tensor field of type (0, 2) in F<sup style="top: -0.3298em;">∗</sup>(M<sub style="top: 0.1364em;">n</sub>) which called the diagonal lift of the tensor field g to F<sup style="top: -0.3298em;">∗</sup>(M<sub style="top: 0.1364em;">n</sub>) with respect to Γ<sup style="top: -0.3299em;">k</sup><sub style="top: 0.2546em;">ij</sub>. From (2.6), (2.7) and (3.1) we prove that <sup style="top: -0.3299em;">D</sup>g has components of the form </p><p></p><ul style="display: flex;"><li style="flex:1">ꢃ</li><li style="flex:1">ꢄ</li></ul><p></p><p>g<sub style="top: 0.1364em;">ij </sub></p><p>0<br>0</p><p><sup style="top: -0.3753em;">D</sup>g = </p><p>(3.2) </p><p>δ<sub style="top: 0.1482em;">αβ</sub>g<sup style="top: -0.3298em;">ij </sup></p><p>with respect to the adapted frame {D<sub style="top: 0.1407em;">I</sub>} in F<sup style="top: -0.3298em;">∗</sup>(M<sub style="top: 0.1364em;">n</sub>) and components </p><p></p><ul style="display: flex;"><li style="flex:1"> </li><li style="flex:1">!</li></ul><p>P</p><p>g<sub style="top: 0.1363em;">ij </sub>+ <sub style="top: 0.2688em;">α=1 </sub>g<sup style="top: -0.3299em;">ks</sup>X<sub style="top: 0.2247em;">m</sub><sup style="top: -0.3299em;">α </sup>X<sub style="top: 0.2767em;">l</sub><sup style="top: -0.3299em;">α</sup>Γ<sub style="top: 0.2767em;">i</sub><sup style="top: -0.3299em;">m</sup><sub style="top: 0.2767em;">k</sub>Γ<sub style="top: 0.2546em;">j</sub><sup style="top: -0.3299em;">l </sup><sub style="top: 0.2546em;">s </sub>−g<sup style="top: -0.3299em;">js</sup>X<sup style="top: -0.4423em;">β</sup>Γ<sub style="top: 0.2546em;">i</sub><sup style="top: -0.3299em;">l</sup><sub style="top: 0.2546em;">s </sub></p><p>n</p><p><sup style="top: -0.3753em;">D</sup>g = </p><p>(3.3) </p><p>l</p><p></p><ul style="display: flex;"><li style="flex:1">−g<sup style="top: -0.3299em;">is</sup>X<sub style="top: 0.2767em;">l</sub><sup style="top: -0.3299em;">α</sup>Γ<sup style="top: -0.3299em;">l</sup><sub style="top: 0.2546em;">js </sub></li><li style="flex:1">δ<sub style="top: 0.1481em;">αβ</sub>g<sup style="top: -0.3299em;">ij </sup></li></ul><p></p><p>with respect to the natural frame {∂<sub style="top: 0.1363em;">i</sub>, ∂<sub style="top: 0.1363em;">i </sub>}, where g<sup style="top: -0.3299em;">ij </sup>denote contravariant com- </p><p>α</p><p>ponents of g. <br>From (3.2) it easily follows that if g is a Riemannian metric in M<sub style="top: 0.1363em;">n</sub>, then <sup style="top: -0.3299em;">D</sup>g is a Riemannian metric in F<sup style="top: -0.3298em;">∗</sup>(M<sub style="top: 0.1364em;">n</sub>). The metric <sup style="top: -0.3298em;">D</sup>g is similar to that of the Riemannian metric studied by Sasaki in tangent bundle T(M<sub style="top: 0.1363em;">n</sub>)[14] (for the cotangent bundle T<sup style="top: -0.3298em;">∗</sup>(M<sub style="top: 0.1364em;">n</sub>)and linear frame bundle F(M<sub style="top: 0.1364em;">n</sub>) see [15],[7], respectively). <br>From (2.1) and (3.3) we have </p><p><sup style="top: -0.3753em;">D</sup>g(<sup style="top: -0.3753em;">H</sup>X, <sup style="top: -0.3753em;">H</sup>Y ) = <sup style="top: -0.3753em;">V </sup>(g(X, Y )) = g(X, Y ) ◦ π. </p><p>(3.4) <br>Therefore we have as follows. </p><p>1</p><p>Theorem 3.1. Let X, Y ∈ =<sub style="top: 0.2436em;">0</sub>(M<sub style="top: 0.1364em;">n</sub>). Then the inner product of the horizontal lifts <sup style="top: -0.3299em;">H</sup>Xand <sup style="top: -0.3299em;">H</sup>Y to F<sup style="top: -0.3299em;">∗</sup>(M<sub style="top: 0.1363em;">n</sub>)with the metric <sup style="top: -0.3299em;">D</sup>g is equal to the vertical lift of the inner product of Xand Y in M<sub style="top: 0.1363em;">n</sub>. </p><p>From (2.1) and (3.3) we have also </p><p>V<sub style="top: 0.1172em;">β </sub></p><p><sup style="top: -0.3753em;">D</sup>g(<sup style="top: -0.3753em;">V </sup>ω, θ) = δ<sup style="top: -0.3753em;">αβV </sup>(g<sup style="top: -0.3753em;">−1</sup>(ω, θ)) = δ<sup style="top: -0.3753em;">αβ</sup>(g<sup style="top: -0.3753em;">−1</sup>(ω, θ) ◦ π), </p><p>(3.5) (3.6) </p><p>α</p><p><sup style="top: -0.3753em;">D</sup>g(<sup style="top: -0.3753em;">H</sup>X, <sup style="top: -0.3753em;">V </sup>θ) = 0 </p><p>β</p><p></p><ul style="display: flex;"><li style="flex:1">1</li><li style="flex:1">0</li></ul><p></p><p>for all X ∈ =<sub style="top: 0.2436em;">0</sub>(M<sub style="top: 0.1364em;">n</sub>) and ω, θ ∈ =<sub style="top: 0.2436em;">1</sub>(M<sub style="top: 0.1364em;">n</sub>). We recall that any element t ∈ =<sub style="top: 0.2435em;">2</sub>(F<sup style="top: -0.3299em;">∗</sup>(M<sub style="top: 0.1363em;">n</sub>)) is completely determined by its action on vector fields of type </p><p>0</p><ul style="display: flex;"><li style="flex:1">DIAGONAL LIFTS OF METRICS TO COFRAME BUNDLE </li><li style="flex:1">331 </li></ul><p></p><p><sup style="top: -0.3299em;">H</sup>X and ω. From this it follows that <sup style="top: -0.3299em;">D</sup>g is completely determined by its eqs </p><p>V<sub style="top: 0.083em;">α </sub></p><p>(3.4), (3.5) and (3.6). </p><p>D</p><p>4. Levi-Civita connection of g </p><p>In section 2 we see that the components of the adapted frame {D<sub style="top: 0.1407em;">I</sub>} are given by </p><p>ꢂ</p><p>(2.2) and (2.3). On the other hand the components of the dual coframe η˜<sup style="top: -0.3299em;">I </sup>are given by matrix (2.6). <br>Since the adapted frame is non-holonomic, we put </p><p>[D<sub style="top: 0.1408em;">I</sub>, D<sub style="top: 0.1408em;">J </sub>] = Ω<sub style="top: 0.2247em;">IJ </sub><sup style="top: -0.3753em;">K</sup>D<sub style="top: 0.1408em;">K </sub></p><p>from which we have </p><p></p><ul style="display: flex;"><li style="flex:1">ꢀ</li><li style="flex:1">ꢁ</li></ul><p></p><p></p><ul style="display: flex;"><li style="flex:1">K</li><li style="flex:1">K</li></ul><p></p><p>Ω<sub style="top: 0.2248em;">IJ </sub>= D<sub style="top: 0.1407em;">I</sub>A<sub style="top: 0.2248em;">J</sub><sup style="top: -0.3754em;">L </sup>− D<sub style="top: 0.1407em;">J </sub>A<sub style="top: 0.2248em;">I</sub><sup style="top: -0.3754em;">L </sup>A <sub style="top: 0.2248em;">L</sub>. </p><p>¯<br>According to (2.2), (2.3), (2.6) and (2.7), the components of non-holonomic object </p><p>K</p><p>Ω<sub style="top: 0.2694em;">IJ </sub>are given by </p><p>(</p><p>k<sub style="top: 0.083em;">γ </sub>k<sub style="top: 0.083em;">γ </sub>k<sub style="top: 0.083em;">γ </sub></p><p></p><ul style="display: flex;"><li style="flex:1">= −Ω<sub style="top: 0.2714em;">j i </sub>= −δ<sub style="top: 0.2935em;">β</sub><sup style="top: -0.4423em;">γ</sup>Γ<sup style="top: -0.4422em;">j</sup><sub style="top: 0.2935em;">ik</sub>, </li><li style="flex:1">Ω<sub style="top: 0.2714em;">ij </sub></li></ul><p>Ω<sub style="top: 0.2714em;">ij </sub></p>
Details
-
File Typepdf
-
Upload Time-
-
Content LanguagesEnglish
-
Upload UserAnonymous/Not logged-in
-
File Pages10 Page
-
File Size-