Sanjay, my experience is different with uni-axially pressed samples after sintering. Which materials you use, maybe the compacting pressure is too high so you have severely expressed regions with different local density?
These deformations are usually due to stress into green pellets. Bending increases with pellet size. One possibility to avoid bending is to fabricate small pellets (diameter < 0.5 cm or thikness < 0.3 cm). An other possibility is to use isostatic compaction or uniaxial pelletization with three part die.
Dear Sanjay, I am interesting by your question as it was nearly my PhD Subject (good memories).
Deformation of uniaxial compacted powder during thermal treatments originates from the low viscosity of not completely sintered materials at the temperature of the treatment. Note that this viscosity strongly evolves with sintering. The origin of the stress inducing bending can be : i) internal stress due to uniaxial compaction and ejection; ii) creep induced by gravity iii) heterogeneous shrinkage due to density gradient iv) anisotropic sintering shrinkage. Then you have to improve the ejection, to support the sample for gravity, the density gradient can be reduced but not completely eliminated, and for anisotropy the only way is to take it into account by modifying the shape of the green part.
during uniaxial pressing of a powder the densification will not be homogeneous, due to friction of the powder at the wall of the form. As consequence the density at the side of the moving Piston will be higher and differs over the hight of the cylinder. This can be improved a bit by two Pistons pressing from both sides, but better will be using a Floating cylinderliner which reduces the effect of the wall friction or even more, as Thibaud mentioned, using isostatic or hot-isostatic compaction.