Failure of castings in the mining industry: applications of fracture mechanics

Transcription

1 Failure of castings in the mining industry: applications of fracture mechanics Dr.Richard Clegg Director, Explicom Pty Ltd Editor-in-Chief, Engineering Failure Analysis Adjunct Professor, Queensland University of Technology

2 Outline Engineering Failure Analysis Failure analysis and fracture mechanics Case studies

3 Engineering Failure Analysis Publisher: Elsevier Published since 1993 Current submissions around 1130 so far in 2018 Acceptance rate ~ 30% Rankings Impact Factor ( ) 53 out of 130 Mechanical Engineering journals (Q2) 11 out of 33 Materials Science, characterisation and testing (Q2) CiteScore ( ) 37 out of 265 journals (Q1)

4 Australia s Top Exports Iron ores, concentrates $49.3 billion 2 Coal, solid fuels made from coal $40.6 billion 3 Petroleum gases $20.5 billion 4 Gold (unwrought) $13.1 billion 5 Aluminum oxide/hydroxide $5.8 billion 6 Wheat $4.7 billion 7 Crude oil $4 billion 8 Copper ores, concentrates $3.6 billion 9 Frozen beef $3.5 billion 10 Wool (uncarded, uncombed) $2.9 billion

5 Costs of failures Failures occur in industry and can lead to significant consequential loss. Major incidents can be costly Varanus Island (June 2008) Deepwater Horizon (2010) Typical Costs of Failure Replacement components Loss of production Loss of confidence in products Injury or death Damage to reputation and social licence to operate Consequential damage (environmental, societal costs) Varanus Island

6 Incident investigations and failure analysis Very few incidents are acts of God or an unexpected combination of circumstances. Many incidents are the result of a process of safety degradation. Past experience has shown that major incidents rarely just happen, but are typically pre-dated by a culture that tolerates small failures. Risk management and safety have become an important part of corporate culture in Australian mining. The investigation of unexplained failures, even minor, has become an important part of the culture of companies.

7 Framework for viewing failures Design Everything that was the responsibility of the design team - OEM Design Use Manufacture Manufacture Everything that was the responsibility of the OEM (including subcontracted manufacturers) in ensuring that the design was correctly realised in a manufactured form Use - Everything that was the responsibility of the end user, including method of operation, repairs and unauthorised modifications

8 Good Failure Analysis Observation Keen observation Reliable background Hypothesis Good experimental work Knowledge and understanding of failures Synthesis Communicate with stakeholders

9 Understanding failures Basic Engineering Science Knowledge and understanding of failures Experienced mentors Communities of practice Manufacturers and suppliers Past Experience Personal Experience Published Case Studies

10 Where can metallurgical examination help! Useful Triage - A metallurgical examination needs to be part of a wider failure analysis Main strengths are in: Help determine the failure story HOW did the failure happen? Identification of the mechanisms of failure: Creep, fatigue, stress corrosion cracking, surface fatigue etc. Generally, this is descriptive and not numerical (fracture mechanics) Confirming the grade of material used Identifying any metallurgical defects present (quality of the manufacture).

11 What is a Failure Analysis - Opinion or Fact? The result of a failure analysis is an OPINION Not a FACT. Scientific and engineering forensic analysis are used to develop and support an OPINION. Experimental work can be used to either support or discount theories.

12 Fracture Mechanics and Failure Analysis Where does fracture mechanics fit into failure analysis? Help improve the failure story - Sanity check. Answer specific questions. Characteristics of mining equipment Large components, high strength. High costs Reliability is important General problems Fatigue, corrosion, wear Some specific problems (mercury embrittlement in gas plants, caustic cracking in alumina refineries)

13 Fatigue failure Large corner crack in the I beam section

14 Fracture Mechanics Fracture mechanics is the field of mechanics concerned with the study of the propagation of cracks in materials. (Wikipedia). Applications Fundamental understanding of crack behaviour Material property determination (quality assurance) Fitness for Service assessments (BS7910, API579, SINTAP etc) Failure analysis Application in Failure Analysis is somewhat limited, despite the potential to provide quantitative assessments of crack-related problems.

15 Fitness for service applications What do I do if I find a crack or defect? Ignore? Repair immediately? Schedule a repair in the near future? What is the largest defect that I can tolerate in my structure? How fast will a defect grow? Structures typically assessed to standards such as AS3788 (Pressure Equipment - In-Service inspection) with reference to design codes such as AS1210 Pressure Vessel Code Tools for structural integrity API579 Fitness for Service BS7910 Guide to methods for assessing the acceptability of flaws in metallic structures SINTAP European guide Naturally need to be conservative design and FFS assessments.

16 Failure analysis (forensic analysis) compared with FFS Forensic Analysis The aim is to provide information to enable the investigation to proceed We know the position and orientation of the failure Operating conditions not necessarily well-known Material can be tested Engineering analysis can be as exact as the data can allow and the client requires Documentation required along with retention of evidence Design Analysis The aim is to provide safe design Engineering judgement required to predict possible failure modes Best estimates of stress and loads are available Material data dependent on data sheets and predictions from manufacturing (CTOD Testing) Design codes provide conservative analysis (e.g. BS7910/API579) Documentation required

17 Failure Analysis Diagram Approach In assessing a failed component, several standards are a useful starting point API579 Fitness for Service BS7910 Guide to methods for assessing the acceptability of flaws in metallic structures SINTAP European guide Generally used for Fitness for Service assessments, but contain useful data on stress intensities and toughness estimates. Can be used as a basis for forensic analysis of fractured components WTH CARE.

18 Fracture of structures failure of load bearing structures generally by yield or fracture Fracture of structures Yeilding dominant General plasticity Defects are (sub)microscopic Fracture - dominant Highly locallised plasticity Defects are macroscopic 18

19 Domains of Fracture Mechanics Linear elastic fracture mechanics Elastic plastic fracture mechanics Plastic collapse Ductility Defect size 19

20 Brittle Fracture - Stress intensity In brittle materials, the magnitudes of the stresses ahead of the crack tip can be fully characterised by a single figure, the stress intensity, K I. KI = a Solution assumes that almost all of the material behaves elastically (small scale yielding only) Deviations from the geometry described can be incorporated in the stress analysis. Usual to incorporate them as a factor, Y in the stress intensity equation KI = Y a 20

21 Crack like flaw assessment API579 Section 9 Level 1 Screening assessment. Consists of plots of allowable flaw length versus temperature Level 2 Basic fracture mechanics analysis. Uses the failure assessment diagram (FAD) approach Level 3 Advanced fracture mechanics analysis. May involve finite element analysis of a component with a crack

22 Level 2 Assessment Use of fracture mechanics to assess criticality of flaws Data required Defect geometry Material properties Yield and tensile strength Fracture toughness Crack growth model (?) Loads/stresses Primary stresses (membrane and bending) Secondary stresses and Residual stress Material Loading Crack

23 Failure Analysis Diagram K r = K I P + Φ 0 K I SR K mat 0.8 Not Acceptable K r Acceptable L p r L r P = σ ref P σ y

24 Prestressing wire failure A client installed prestressing cables to fix a temporary foundation system in a construction. In the hours and days after the cables were fitted, the cables began to fail and eventually the temporary foundation walls collapsed into the construction site. I was asked to look into why the cables failed.

25 Metallurgical Failure Analysis Property Value Tensile strength (MPa) 1780 Proof stress (0.1%) (MPa) 1550 Elongation 5% C Mn P S Si Cr N Failed New Many side cracks could be seen

26 Nature of the factures

27 Microstructural analysis Cold drawn high carbon steel

28 Cause of failure Wires consisted of high strength, cold drawn, high carbon steel. Stress corrosion cracking had led to multiple side cracks SCC cracks were up to 300 microns deep on the sections examined. What is the critical size for the SCC crack for failure? Why do I want to know this?

29 Example High Strength Wire What is the critical size for defects in prestressing wire? Treat wire as a rod with circular cross section and a surface flaw Critical Data Defect geometry unknown Diameter of wire, 5 mm Material properties Yield strength, 1550MPa Fracture toughness, 40MPa m 1 Loads/stresses 1000MPa (membrane only) Note 1. Estimated from literature

30 Determination of Ratios (API579) Reference Stress L P r = σ P ref σ y Stress Intensity K r = K I P +Φ 0 K I SR K mat

31 Evaluation using FAD a = 0.70 mm a = 0.45 mm a = 0.25 mm a = 0.05 mm

32 Higher stress (1250 MPa) a = 0.45 mm a = 0.25 mm a = 0.05 mm

33 Lower Stress (750 MPa) a = 0.75 mm a = 0.45 mm a = 0.25 mm a = 0.05 mm

34 Estimation of fracture toughness Fracture toughness is a critical parameter and not commonly measured. Estimates often need to be done on limited material and at low cost. Methods for estimating toughness Direct measurement ASTM E399/E1820/E1921 Costly May need more material than is available. Handbook/literature data Not necessarily relevant to application Embrittled material may not be well characterised Indirect measurement Charpy Instrumented Charpy (notched vs precracked) Theory of Critical Distances

35 Indirect measurements Charpy Data - Several Attempts at correlating Charpy data to toughness summarised in API 579 and BS Usually correlations are dependent on position on transition curve - Lower Shelf - Transition - Upper Shelf - Often rely on the establishment of the Reference temperature, T 0. K Jc = exp T T 0 - Methodologies generally determine Lower Bound toughness values whereas for failure analysis purposes, median values are probably more important

36 Instrumented Charpy Data Instrumented Charpy machines can provide greater information than Charpy (load vs displacement data). Some authors propose methods for calculating K jc from notched-only instrumented data (Schindler) ISO Standard concerning instrumented Charpy (ISO 26843). Requires fatigue precracked samples. Testing Authority Test temperature Average Estimate K IC Minimum Charpy Estimate (minimum) Charpy (average) (J) (MPa m) (J) (MPa m) Manufacturer Ambient Testing Auth1-15 C Testing Auth2-15 C K IC

37 Instrumented Charpy data

38 Load (N), Cumulative Absorbed Energy (J) EDM notched samples at 21 C. 20 Force Sample A Force Sample T Force Sample G 15 Force Sample M Force Sample 11 Energy Sample A Energy Sample T 10 Energy Sample G Energy Sample M Energy Sample Displacement (mm)

39 Load (N), Cumulative Absorbed Energy (J) EDM Notched at -20C. 20 Force Sample 2 Force Sample 5 Force Sample 7 Force Sample L Force Sample 11 Energy sample 2 Energy Sample 5 Energy Sample 7 Energy Sample L Energy Sample Displacement (mm)

40 EDM Notched samples Test Temperatur e Sample No K I N (See 1) Note dk IN /dt (See 2) Note R pd (See 3) Note J d (See 4) Note K Jd (See 5) ( C) (MPa m) (MPa m/s) (MPa) (J/m 2 ) (MPa m) Note Type Force Diagram of Validity Criteria (See 6) Note , I Yes , I Yes , I Yes -20 L , I Yes , I Yes Average , , , I Yes , I Yes , I Yes , I Yes 0 J , I Yes Average , , dj d /dt (J/m 2 /s) 21 T e e II No 21 A e e II No 21 G e e II No 21 M e e II No e e II No Average e e

41 Load (N) Fatigue Pre-cracked at -15C Time (s)

42 Fatigue pre-cracked Instrumented Charpy Pros Relatively small amount of material Easy to test at different temperatures Standard equipment (?) Cons Must fatigue pre-crack Small size Standard is a little bit complex and is limited in scope Temperature ( C) K Jc (MPa m) K Jc (MPa m) (1T) ambient

43 Pump crankshaft Failure of several large cast crankshafts (7 tonne) Failure by fatigue at journal radii.

44 Fracture surface

45 Casting porosity

46 Crack initiation

47 da/dn (mm/cycle) Fatigue curve 1.00E At R=0.1, Kth = 7 MPa m 1.00E E E E-07 Stress Intensity Range (MPa m) R=0.1 Lower Upper R=0.5

48 Stress analysis of crankshaft FEA Analysis showed that maximum cyclic stresses were around 230 MPa. Approximate stress intensity range for 8 mm deep crack K= 40 MPa m What is the defect tolerance? For K = 7 MPa m, a<0.25 mm

49 Summary Failure analysis continues to be an important process for assuring quality and safety Fracture mechanics can play a part, but needs to be: Reasonably accurate Streamline and efficient Cheap to implement Fracture mechanics can be used to improve opinions on the causes of failures.