report

A

B

DUCTILE FAILURE

(parrot nose fracture)

BRITTLE FAILURE

Resin B

Catalytic

particles

Variability

For each size there should be at

least 20 valid data points

Reject all data with a failure in less

than 10 hours or any surviving pipe

No more than 2 failures in less

than 100 hours

British gas specs:

Average should be better than 50 yrs

at 8.3 MPa

95% probability of survival at 8MPa

and 100,000 hours (11.4 yrs)

Ductile

Brittle

NO DATA !

Be careful what behavior

is extrapolated!....

•At a given stress, find time to failure for different temperatures

•Plot time to failure versus reciprocal of temperature

•Extrapolate TO service temperature

MDPE better at long service times !....

Ductile Brittle

Environmental Stress-Cracking Resistance (ESCR) - ASTM D1693

Three test conditions are specified. Condition A is generally used for polyethylene

with densities between 0.910 and 0.925 g/cm³. Condition B is used for polyethylene

with densities greater than 0.925 g/cm³. Condition C is used for accelerated testing of

materials with extremely high ESCR values.

A summary of the differences in the different test conditions is listed below.

ESCR Test Conditions

Condition Thickness (mm) Notch Depth (mm) Bath Temperature (°C)

A min 3 0.5

50 max 3.3 0.65

B min 1.84 0.3

50 max 1.97 0.4

C min 1.84 0.3

100 max 1.97 0.4

Time to failure

A major problem:

Do differences among times- to-failure mirror a real difference in

ESCR, or do they merely reflect the higher stress levels in the

stiffer specimens?

Crack growth velocity

(defect tolerant approach)

Specimen shape

Rate dependence of yield stress

Thus the plastic zone is rate dependant!...

Plane strain/Small scale yielding

Bmin=2.5 (K/Y) 2 < specimen thickness

2.5 (K/Y) 2 is about 15 rYC (rYC =plastic zone size at fracture)

Also recall that at these very thick samples residual stresses and non-

homogeneous microstructure make things complicated…

• Therefore fracture

toughness is a

function of thickness.

Runaway crack.

• Assume that we can get an “estimate” of KIC

If a crack advances by da, a new crack surface of 2 t da is created and a piece of the pipe is relieved of the stored strain energy

da

datDV    

  22

22

Stored energy – released as crack advances

Energy release

2

2

2 22

2

D K

E

K D

dat

datD

G

IC

IC IC

 

 

 









 



2

D K

IC

 

Therefore – the larger pipes are more susceptible to

rapid crack growth….

Criteria

• hoop strain of less than 3% after 50 years

• no short term collapse due to load of 20kN/m-2

• Small change of diameter due to creep under soil pressure

• Average life at 50 years or more at 8.3 MPa

• a 95% probability that the lifetime exceeds 100,000 hours (11.4 years) at 8 MPa

• “Good” environmental stress cracking resistance

• No rapid crack propagation

• Low cost

Operating pressure 2 bar

Note that all the above apply to ALL sizes…