You can only calculate the theoretical density of the crystalline component using the nit cell data of a single crystal X-ray measurement But actually, it is much easier to measure the density directly.
I absolutely agree with your note with respect to the impossibility of producing 100% crystalline polymer, BUT single crystals make possible the determination of unit cell structure which gives you a volume and the atomic masses give you a weight, thus you can calculate a density, which gives you an upper limit for the crystalline polymer.
Yes you all are right that polymers are not 100% crystalline. But I was talking about conducting polymers which are semicrystalline in nature. Rather we calculate the density with using Archimedes principle but we want to compare it with XRD data.
Is it possible to prepare quasi-single crystal from your sample? (Comparable in quality with e.g. PE single crstals, which are really folded lamellae). If yes, do you know the 3D crystal structure of your polymer?
If your intention is to compare density calculated by Archimedes principle with XRD data, what you actually wanted to calculate is the relative density of your samples from both theoretical and experimental densities. Theoretical density can be achieved by Archimedes principle while Experimental density from XRD.