Hi,

how do you mean the segments?  Lengthwise, radialwise? Best would be to make a simple drawing, even in MS Paint.

Thanks for the drawing. Wished to know how you meant.

I do not think it would give any advantage, especially not if there are air gaps left between the segments.

In electromagnets, the core's permeability and the number of turns that mainly counts. Attract force depends on B² so the core should have high saturation flux data, and you have to consider the cross section area of the core. 
And if you have to switch the electromagnet on and off with a certain frequency (even if with only 5-10 Hz), then you have to start to consider eddy current losses in the core.

Have you seen this post by member DMMPOWER here:
http://www.overunity.com/index.php?topic=4624.msg96814#msg96814

He sums the flux of the permanent magnet with the flux of the electromagnet, so a stronger electromagnet is the result for the same input excitation. Be careful to choose the core material to avoid saturation from the summed fluxes. And a good sing is that when the coil is off the flux of the permanent magnet is closed within the core, not much or no leakage to the outside.

If you use the Search function on this Forum with key word electromagnet, you may find further hints on them.
 Gyula

Well, now that you asked that, I am a bit puzzled because I have not done any tests with segmented core electromagnets.  Segmentation has never been an issue with cores,  I am not aware of practical solutions applying segmentation in electromagnet cores. 
All I can logically think on possible drawback, hence possible decrease in magnetic field, is the air gap between the segments that impose much higher reluctance for the flux than the core itself.  And if you do your best to match the facing surfaces of the segments to reduce the air gap than you pretty much approach the single solid core case, with no segmentation...

I wonder why using segments for a core is needed?