Naked Science Forum

Non Life Sciences => Physics, Astronomy & Cosmology => Topic started by: Richard777 on 21/07/2016 00:05:14

Title: Does a Tensor have "magnitude"?
Post by: Richard777 on 21/07/2016 00:05:14
A tensor may be represented as a product of vectors (arithmetic product, not cross, not dot), such as a four vector.
A vector has a magnitude (scalar).
May a tensor have a magnitude?
May it be represented as the product of vector magnitudes?
If so this may lead to interesting physics.

Title: Re: Does a Tensor have "magnitude"?
Post by: PmbPhy on 21/07/2016 17:02:25
Quote from: Richard777
A tensor may be represented as a product of vectors (arithmetic product, not cross, not dot), such as a four vector.
Where did you get that assumption from? The fact that the product of vectors is a tensor doesn't mean that the converse is true. For example; The Faraday tensor, aka the electromagnetic field tensor, is a second rank tensor. This tensor cannot be written as the product of two 4-vectors.

Quote from: Richard777
A vector has a magnitude (scalar).
May a tensor have a magnitude?
Not in general. You can always contract a tensor until all that's left is a scalar (i.e. a tensor of rank 0) but that won't yield a quantity that is very descriptive of the what the tensor is defined as.

Note: Unlike your high school math and science classes, the term scalar is defined as a tensor of rank zero. For reference please see: http://mathworld.wolfram.com/Scalar.html

Quote from: Richard777
May it be represented as the product of vector magnitudes?
If so this may lead to interesting physics.
If you have a tensor which is defined as the product of 4-vectors then you could define the magnitude as the product of the norms of the 4-vectors. But again, this may not be useful. For example; consider 10 tensors, each of which defined in terms of a number of 4-vectors, each tensor having a different number of 4-vectors. Let each 4-vector be the 4-momentum of a photon of various energies, each photon having a different energy. Then regardless of the number of 4-vectors in each tensor, the magnitude so defined will always be zero since the norm of a photons 4-momentum is zero.

Question: Do you want the magnitude to be invariant, i.e. a scalar quantity?