-
1
- #1
dorios
Mechanical
- Feb 8, 2004
- 8
Hi,
** this is a very specialised subject for people skilled in the design of mitre bends***
we conducted extensive FEA to validate a number of mitre bends (90 deg , 4 segment bends, theta=30).
asme b31.3 section 304.2.3 gives two formulae: one (eq 4a) which is independent of segment A (=M/2)and another one (eq 4b) which is dependent of segment A
our FEA show that (eq 4a) is too conservative but the Code says to use the "lesser value" of the two formulae.
For instance for an OD1400mm, t=11mm, A=400 mm (M=800 mm), formula 4a gives an internal pressure of 1.1 MPa, formula 4b gives an internal allowable pressure of 1.33 MPa while our FEA shows that stress at intrados (tresca stress) reaches yield at around 1.25-1.3 MPa.
although these figures are close for A=400 mm, when A is increased (to say to A=1050 mm in other words there is a much larger curvature, but keep OD and t constant) eq 4a recommends same P=1.1 MPa while eq 4b gives me 1.65 MPa and FEA around 1.75 MPa (based on a max stress (tresca) of 250 MPa very localised on a v small area of intrados, while the rest of the bend and bend legs is at a safe value of around 117 MPa.
I agree with formula 4b which is and should be A dependent.
AWWA C208 offical notice for the design of large water bends also gives a formula which is A dependent and follows closely in shape and value eq 4b in asme b31.3
formula 4a is a flat line if plotted on a graph of working pressure vs A while formula 4b shows a curve that increases working prtessure as A increases... which is normal as with the increase in A (maintaining t and OD constant) R1 increases which makes a larger curvature, less stress concentrators at intrados.
does anyone have experirnce in designing large pipe mitre bends (OD500 mm and above)?
thanks
dorios
** this is a very specialised subject for people skilled in the design of mitre bends***
we conducted extensive FEA to validate a number of mitre bends (90 deg , 4 segment bends, theta=30).
asme b31.3 section 304.2.3 gives two formulae: one (eq 4a) which is independent of segment A (=M/2)and another one (eq 4b) which is dependent of segment A
our FEA show that (eq 4a) is too conservative but the Code says to use the "lesser value" of the two formulae.
For instance for an OD1400mm, t=11mm, A=400 mm (M=800 mm), formula 4a gives an internal pressure of 1.1 MPa, formula 4b gives an internal allowable pressure of 1.33 MPa while our FEA shows that stress at intrados (tresca stress) reaches yield at around 1.25-1.3 MPa.
although these figures are close for A=400 mm, when A is increased (to say to A=1050 mm in other words there is a much larger curvature, but keep OD and t constant) eq 4a recommends same P=1.1 MPa while eq 4b gives me 1.65 MPa and FEA around 1.75 MPa (based on a max stress (tresca) of 250 MPa very localised on a v small area of intrados, while the rest of the bend and bend legs is at a safe value of around 117 MPa.
I agree with formula 4b which is and should be A dependent.
AWWA C208 offical notice for the design of large water bends also gives a formula which is A dependent and follows closely in shape and value eq 4b in asme b31.3
formula 4a is a flat line if plotted on a graph of working pressure vs A while formula 4b shows a curve that increases working prtessure as A increases... which is normal as with the increase in A (maintaining t and OD constant) R1 increases which makes a larger curvature, less stress concentrators at intrados.
does anyone have experirnce in designing large pipe mitre bends (OD500 mm and above)?
thanks
dorios