Consider a sufficiently nice topological space $X$ as well as topological groups $G$ and $H$. Consider the functor $F$ that associates to $X$ the set of all isomorphism classes of all principal $H$-bundles $\pi_H\colon P_H\twoheadrightarrow P_G$ on all principal $G$-bundles $\pi_G\colon P_G\twoheadrightarrow X$ on $X$. I think that this would be representable by the Brown representability theorem, i.e. there is a topological space $C_{G,H}$ such that $[X,C_{G,H}]$ is in bijection with the isomorphism classes of such bundles-over-bundles (modulo pointed-vs-unpointed spaces). But is there an explicit description/characterisation of it?

I am specifically interested in the case where $G=\operatorname U(1)^n$ is a torus (and $H$ is compact Lie); the motivation is that this would describe the so-called Kaluza–Klein reductions of gauge fields on nontrivially fibred torus compactifications.

As a special case, the subset of $F(X)$ that corresponds to $H$-bundles on the trivial $n$-torus bundle $\operatorname U(1)^n\times X$ is classified by the $n$-fold iterated free loop space $\mathcal L^n\mathrm BG$; and in the case $n=1$ this object is well studied. But I don't know how to generalise to arbitary principal torus bundles.

New contributor
Zislu R. is a new contributor to this site. Take care in asking for clarification, commenting, and answering. Check out our Code of Conduct.
  • $\begingroup$ Your notion of "isomorphism classes" is what one would guess? $\endgroup$ Oct 13 at 19:01
  • $\begingroup$ @KonradWaldorf: yes, the obvious notion (in the unpointed case). $\endgroup$
    – Zislu R.
    Oct 13 at 19:26

If I understand the question right, I think the classifying space can be described like this. Let $Map(G,BH)$ be the space of all continuous maps from $G$ to $BH$. Make the group $G$ act continuously on that function space (using the usual free transitive action of $G$ on the space $G$), and form the associated bundle over $BG$: $$ EG\times_G Map(G,BH). $$

  • 1
    $\begingroup$ This looks good, I was just about to write this. $\endgroup$ Oct 13 at 19:06

Your Answer

Zislu R. is a new contributor. Be nice, and check out our Code of Conduct.

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.