You may indeed be able to see H-bonding of glycerol molecules in condensed phases (solution or solid). They should show as shifts of the position of some IR bands (those implied in H-bonding) with respect to the position of these same IR bands in the spectrum of the isolated glycerol molecule, i.e. in the gas phase, or isolated in a rare gas matrix,... where there is no intermolecular H-bonds.
The attached file may be of help, showing the example of VOCs adsorption on ice, and the formation of surface H-bonds with surface FTIR spectroscopy.
I think we should get a broad peak for H bonding at about 3300 cm-1 in concentrated solution. If solution is diluted there will be appearance of additional sharp peak very closed to the broad peak. Intensity of the broad peak will reduce and that of the sharp peak will increase dilution.