I'm trying to normalize the data from multiple studies performed using same platform.

ibrary(gcrma)
library(limma)
downloadedAffyFiles <- list.files(path = "GSE53454_RAW/", pattern = "CEL.gz$",full.names=TRUE)
AffyData <- ReadAffy(filenames = downloadedAffyFiles)

eset <- gcrma(AffyData)

Could someone help on how to access the contents of the expression set?

exprs(eset) gives the log-expression values. 

How do we access the contents like cell type description? For Instance, while working with the series matrix files, pData(eset)[[1]] works.But, I am not able to use the same commands here.

Many thanks

This has nothing to do with limma, and using the bioconductor tag for questions on the Bioconductor support site is a tautology. In future you should think a bit more about the tags you use, as many of the support site helpers only follow a subset of the tags, and you are by definition asking Gordon Smyth and Aaron Lun questions about GEOquery, which is not their package (and you aren't using limma to normalize anyway).

In addition you are going to have to become more self reliant at some point. You have already asked numerous questions about getting the processed data from GEO, which does come with the phenodata you want (IIRC, some of those questions were specifically about the phenotypic data). So you already (should) know that you can get the phenotypic data that way. Now you are getting the raw data and finding that you don't get the phenotypic data. But you already know how to get the phenotypic data, so why ask here instead of just going and getting it?