Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

In Need of Simple Example

  1. Jun 26, 2015 #1
    I am have been searching for the of a tensor product of two vectors, but found seemingly conflicting definitions. For example, one source definition was, roughly, that the tensor product of two vectors was another column vector in a higher dimensional space, and another defined the tensor product of two vectors as resulting in a matrix.

    So, which of the two is correct?
     
  2. jcsd
  3. Jun 26, 2015 #2

    WWGD

    User Avatar
    Science Advisor
    Gold Member

    Maybe the issue is one of dualization, but the product of a k-tensor T and an n-tensor S is the (k+n)-tensor given by ##T\otimes S:= T(x_1,..,x_k) \otimes S(x_{k+1},..,x_{k+n})## EDIT: Here I assume vectors are treated as 0-tensors.
     
    Last edited: Jun 26, 2015
  4. Jun 26, 2015 #3

    micromass

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2016 Award

    You will need to show those sources.
     
  5. Jun 27, 2015 #4

    HallsofIvy

    User Avatar
    Staff Emeritus
    Science Advisor

    I have never heard of this. What was the source? Perhaps you are confusing a tensor product with a direct product.

    I doubt you read this correctly. The tensor product of two vectors (i.e. two first order tensors) is a second order tensor tensor which in a given coordinate system can be represented by a matrix. You should be careful to distinguish between those two concepts. Physically, velocity is a vector. But a velocity vector with speed v can be represented by the array (v, 0, 0) in one coordinate system, (0, v, 0) in another, and generally (a, b, c) with [itex]a^2+ b^2+ c^2= v^2[/itex]. A tensor can be represented as an array of numbers in a specific coordinate system.

    Strictly speaking, neither is correct! But the second is a little closer.
     
  6. Jun 27, 2015 #5

    WWGD

    User Avatar
    Science Advisor
    Gold Member

    Maybe if you had a natural isomorphism ##V \rightarrow V^{*} ## (e.g., having an associated non-degenerate form ), you can see your vectors naturally/canonically as linear maps and then you get the standard definition ## v\otimes w \approx. (T\otimes S)(x,y):=T(x)S(y)##.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: In Need of Simple Example
Loading...