PSI - Issue 11
P. Bamonte et al. / Procedia Structural Integrity 11 (2018) 322–330
328
7
P.
Bamonte and A.
Taliercio/ Structu
ral Integrity Pro
cedia 00 (2018)
000–000
the t pred
wo approach icted by the g
es is more th lobal and the
an satisfactor reduced mod
y. The same el (Fig. 7b).
consideration
applies as r
egards the tem
perature pro
files
(a) Fig. 5. Normal s
(b) ravity; (b) wind.
tresses in the lon
gitudinal directio
n caused by the
static loads: (a) g
(a) ig. 6. Temperatu ear elastic be s the hoop s of vertical s aterial streng linear field w s analyzed. failure criter thermal loa site applies in ck pattern to sile strength eveloping jus g to the cont whereas a str ar the vertica ach the tensil e propagatio
(b ver different sect stributions of ted accordin lent (Fig. 8b). nsion and co t by means of ented in Abaq hoop stresses art. Using a hus, the mod r the upper ottom of the aximum princ tion is visibl cracks activ Pa) faster tha al crack.
) ions: (a) z = 0; (b the thermal h g the two m The stress va mpression: th the XFEM. us: according increase fas single, homog el was subdi part, 1 MPa crown. These ipal stress, a e at the crack ate at approx n hoop stress
(c) ) z = 20 m; (c) z oop and vert odels is som lues compute is is why an To reduce the ly, an exped ter than vert eneous, isotr vided in two for the lowe cracks activa stress relief in tip. Fig.s 9c imately 21% es (1 MPa). T
(d)
F
re distribution o
= 40 m; (d) z = 6
0 m.
A
ssuming a lin ined. Wherea ement in term above the m ts in the non ced model wa o orthotropic ation. As the eas the oppo the real cra different ten ontal crack d sis. Accordin be observed, cal cracks ne cal stresses re d highlights th
havior, the di stress compu tress is excel th, both in te as carried ou ion is implem d increases, the upper p be matched. T s (0.2 MPa fo t below the b ours of the m ess concentra l ribs. These e limit (0.2 M n of the vertic
ical stresses s ewhat differ d according t in-depth stu computation ient is adopted ical stresses opic material domains cons r part). Fig.s te at approxim the areas adj and d show of the analys he compariso
hown in Fig. ent (Fig. 8a) o either mode dy of the the al effort, onl to overcome in the lower model woul isting of mate 9a and b sh ately 10% o acent to the c the occurren is, indicating n between Fi
8 are , the l are rmal y the this part, d not rials ow a f the rack ce of that g. 9c
obta agre well effec redu N limit wher allow with horiz analy can verti verti and
Made with FlippingBook Annual report maker