Trivial fiber bundle vs product space

Click For Summary

Discussion Overview

The discussion revolves around the relationship between trivial fiber bundles and product spaces in topology, exploring definitions, properties, and implications in mathematical and physical contexts. Participants analyze the distinctions and similarities between these concepts, particularly in terms of canonical projections and diffeomorphism.

Discussion Character

  • Debate/contested
  • Technical explanation
  • Conceptual clarification

Main Points Raised

  • Some participants propose that a trivial fiber bundle consists of distinct copies of the fiber space, while a product space involves a single copy of the fiber that is identified across the product.
  • Others question the definitions of trivial bundles, referencing external sources like Wikipedia and Penrose's work to clarify their understanding.
  • One participant distinguishes between a trivial bundle and a globally trivializable bundle, suggesting that the former is a product while the latter is homeomorphic to a product without a canonical homeomorphism.
  • Some argue that a product has a richer structure due to its canonical projections, while others contend that if a bundle is diffeomorphic to a product, it retains the same structure.
  • There is a discussion about the nature of canonical projections, with some asserting that a product has two natural projections, while others dispute the notion of what is considered canonical.
  • Participants express differing views on whether the choice of diffeomorphism affects the canonical nature of the projections in trivial bundles versus product spaces.

Areas of Agreement / Disagreement

Participants do not reach a consensus on the definitions and implications of trivial fiber bundles versus product spaces. Multiple competing views remain regarding the nature of canonical projections and the significance of diffeomorphism.

Contextual Notes

Limitations in definitions and interpretations are evident, with participants referencing various sources and examples that may not align perfectly. The discussion highlights the complexity of the concepts involved and the nuances in terminology.

  • #31
jbergman said:
Not following all the arguments that closely but I think I agree with Martin. Once you learn some category theory, products are defined as something satisfying a specific universal property and are unique up to isomorphism. It's fairly easy to construct "different" products that satisfy the dame universal property.
Let me try to recap my understanding about this. In the context of category theory, given two objects ##X_1## and ##X_2## of a given category (say topological spaces), one can define/construct different products. A product consists of two pieces of information: an object ##X## (of the same category) and a pair of projections ##\pi_1: X \to X_1##, ##\pi_2: X \to X_2## which satisfy the universal property. Such a product (i.e. object plus projections named canonical projections) is unique up to canonical isomorphism.

What is such canonical isomorphism ?

Say we have an object ##X'## plus two projections ##\pi'_1: X' \to X_1##, ##\pi'_2: X' \to X_2## satisfying the universal property, i.e. it is a product. Then there exist an unique (canonical) isomorphism ##\varphi: X \to X'## such that ##\pi'_1 = \pi_1 \circ \varphi## and ##\pi'_2 = \pi_2 \circ \varphi##.
 
Last edited:
Physics news on Phys.org
  • #32
Here is a definition from Milnor's Characterisitc Classes , one of the most important texts in 20th century topology.

"
DEFINITION. A real vector bundle § over B consists of the following:
1) a topological space E= E(§) called the total space,
2) a (continuous) map p: E →B called the projection map, and
3) for each be B the structure of a vector space* over the real numbers in the set p^(-1)( b ) .

These must satisfy the following restriction:

Condition of local triviality. For each point b of B there should
exist
a neighborhood UC B, an integer n>0, and a homeomorphism :

h UX R^n →p^(-1)(U)

so that, for each b in U, the correspondence x →h(b, x) defines an isomorphism between the vector space R^n and the vector space p^(-1) (b).

Such a pair (U,h) will be called a local coordinate system for § about b If it is possible to choose U equal to the entire base space, then § will be called a trivial bundle." pg 13 (italics added)

In the category of vector bundles and vector bundle morphisms, triviality is a structural feature. This is why Milnor says, "If it is possible to choose U equal to the entire base space, then § will be called a trivial bundle."
 
Last edited:

Similar threads

  • · Replies 43 ·
2
Replies
43
Views
6K
  • · Replies 13 ·
Replies
13
Views
4K
  • · Replies 15 ·
Replies
15
Views
5K
  • · Replies 29 ·
Replies
29
Views
3K
  • · Replies 5 ·
Replies
5
Views
1K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 10 ·
Replies
10
Views
4K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 6 ·
Replies
6
Views
3K