Besides the fracture mechanism (which is the first issue to analyze here) there is other facts which can influence such behavior, like: blending parameters, compatibilizer, thermal history, ESC, sample architecture, molding process. Please elaborate.

i describe them. i blend PMMA with g-ABS at internal mixer at 190 C then molding at 210 C at hot press with 8min preheat and 10 min hot press. my materiall show ductile behavior at fracture test and crack grows was stable. but  SEM result is very unusual for me and i cant justify this photo

can you justify?

It seems like a co-continus phase, which can happen because of a lack of compatibilizer or bad mixing. Processing seems to be quite Ok. What about the viscosity? A very different MFR is likely to cause bad mixing.

what is the MFR? load -displacement shows good result of toughening and high deformation. if bad mix happened my crack was be unstable or show bad mechanical properties. 

MFR= Melt Flow Rate. I suggest to obtain some mechanical property from raw materials and blend(s), further comparison can tell a lot regarding compatibility/blending.

it is not surprising to me and actually it is sort of similar to what is observed for oxides glasses. A surface created by subcritical crack propagation is smooth (nm to sub nm roughness) where as a fast crack propagation surface looks rougher (it is also known as the mirror-mist-hackle regions). Your stable crack propagation surface seems to be smooth but by having a closer look at it (by atomique force microscopy) you may be able to see the reminisence of plastic events. Regarding the fracture surface generated by unstable crack propagation the roughness is often generated by the occurence of mixed mode resulting from instabilities coming, for example, from the intereraction with the crack tip  of elastic waves emanated from the crack front and boncing back at the sample's boundaries. Also, in order to dissipate as much energy as possible crack branching mays also occur which indeed will create a rougher surface.

I hope this will help

i have a fig with high zoom with stable crack grows can you justify this for me i attach this fig

Whne you brake your polymer in liquid nitrogen , you should be well benath the glassy transition temperature. This could explain the brittle rfacture .

In brittle fracture (isotropic medium), assume the crack to be non-planar (say having a segmented crack-front in planes perpendicular to the direction of fracture propagation); as a result, the magnitude of the average crack extension force (crack extension force averaged over the length of the oscillatory crack front) increases under mixed mode loadings (=v0 - v2 / 3 in pure mode I lower than =v0 + v2 in pure mode II lower than = v0 + 3v2 in pure mode III; v0 and v2 positive quantities: see our contributions below). Consequently, by adopting a non-planar configuration (consisting of planar facets) with an oscillatory crack front in planes perpendicular to the direction of fracture propagation, the propagating crack improves its motion. We may say that the non-planar configuration is the natural configuration of the propagating crack under general loading.

Please go through works dealing with calculation of the crack extension force when the fracture surface is non planar; see our contributions: A STUDY OF THE MIXED MODE I+III LOADING OF A NON- PLANAR CRACK USING INFINITESIMAL DISLOCATIONS ; An analysis of a non-planar crack under mixed mode I+III loading using infinitesimal dislocations with edge and screw average characters ; An analysis of a non-planar crack under general loading using continuously distributed sinusoidal edge and screw dislocations (ResearchGate).

Second, when dislocations are emitted from the tip of the crack, the non-planar configuration of the broken surface is preserved but the shape may be modified.