90deg Bend with shell element

[aigbeawen]

Hi, everyone,
I modelled a 90 deg bend using shell elements, sc8r but the analysis wont converge even at the initial,Please can any one tell me what is wrong.

see the model below:

*HEADING
shell test
*PREPRINT,ECHO=YES,MODEL=NO,HISTORY=NO
*SYSTEM
0.,1.778,0.,0.,1.778,1.
0.,2.778,0.
*NODE, SYSTEM=C
1 ,0.1508,0.,1.778,1.,0.,0.
37,0.1508,360.,1.778,1.,0.,0.
38,0.1778,0.,1.778,1.,0.,0.
74,0.1778,360.,1.778,1.,0.,0.
*Nset, nset=side_d, generate
741,814,1
*NGEN,LINE=C,Nset=n_inner
1,37,1,,0.,0.,1.778,0.,0.,2.778
*NGEN,LINE=C,Nset=n_outer
38,74,1,,0.,0.,1.778,0.,0.,2.778
*Nset,Nset=side_a
n_inner,n_outer
*system
*Ncopy,oldset=side_a, shift, new set=side_b, change number=370
1.778, 0., 0.
0.,0.,0.,0.,0.,0.,0.
*Nfill, Nset=section_1
side_a,side_b,5,74
*Ncopy,oldset=side_b, shift, new set=side_c, change number=74, multiple=5
0., 0., 0.
0.,0.,0.,0.,0.,-1.,18.0
*Ncopy,oldset=side_d, shift, new set=side_e, change number=370
0., -1.778, 0.
0.,0.,0.,0.,0.,0.,0.
*Nfill, Nset=section_3
side_d,side_e,5,74
*Element, type=sc8r
1,2,1,75,76,39,38,112,113
*Elgen, elset=inner_el
1,15,74,1,36,1,15,1
*Shell section, material=steel, elset=inner_el
0.027
*surface, name=in_pres_surf, type=element
inner_el, s1
*surface, name=out_pres_surf, type=element
inner_el, s2
*****
*MATERIAL,NAME=steel
*density
7850.
*ELASTIC
2.07e11,0.3
*PLASTIC,HARDENING=ISOTROPIC
4.48e+08
*boundary
side_a, encastre
side_b, encastre
*step, nlgeom
*static
0.5,1.,1e-05,1.
*dsload
in_pres_surf,P,0.3016,-13.2e6
*OUTPUT,FIELD
*ELEMENT OUTPUT,ELSET=inner_el
S,E
*end step