Finding Errors in Structural Designs

1. James Hanson, Ph.D., P.E. Rose-Hulman Institute of Technology Voted #1 Eight Consecutive Years Finding Errors in Structural Designs How is it Typically Done? ASCE Illinois Section – Structural Group Meeting November 15, 2006

2. Overview • Categories of Errors • Experience of a Generation • Shared Experiences • Tools for Finding Errors • Teaching Tools for Finding Errors

3. Categories of Errors • Idealization of Reality • Assumptions Inherent to Analysis or Design Method • Roundoff Error • Human Error

4. Idealization of Reality

5. Assumptions Inherent to Analysis or Design Method Shear Strength Code Design Eqtn Concrete Beam Depth

6. Roundoff Error

7. Human Error From AISC Seminar “Field Fixes – Common Problems in Design, Fabrication and Erection – Solutions and Prevention”

8. Overview • Categories of Errors • Experience of a Generation • Shared Experiences • Tools for Finding Errors • Teaching Tools for Finding Errors

9. End of a Generation • Precomputer engineers adept at finding errors • Often called “experience” • Will experience be lost?

10. Gathering Experience • Ten Firms: 1 - 700 engineers • Interviewees: 35 structural engr • Experience: 1-55 years • PE’s: 29 • BS: 8, MS: 26, PhD: 1

11. Interviews Critical Incident 1: “Think of the most recent time you discovered something unreasonable in the results of analysis or design.” Critical Incident 2: “Think of the most alarming time you discovered something unreasonable in the results of analysis or design.”

12. Overview • Categories of Errors • Experience of a Generation • Shared Experiences • Tools for Finding Errors • Teaching Tools for Finding Errors

13. Post-tensioned slab-on-grade cracks. Followed standard practice. Standard practice assumptions were not valid in this case.

14. Beam through stairwell at chest height. Accidentally left on drawings. Designer failed to check drawings.

15. Detail not appropriate. Copied from previous design. Reviewer recognized.

16. Beams meet without supporting column. Conditions changed requiring removal of planned column, but beams not redesigned. Found when checking load paths. C B CB CB A 1 3 4 2

17. Using in-house spreadsheet. Changing member length had no effect on design strength. Spreadsheet had an error.

18. Beam flange not thick enough to weld studs. Beam web too shallow to bolt to girder. Designed by computer for strength. Found by experienced reviewer.

19. w Depth (in) wL 2 wL 2 Span (ft) Span (ft) 2 ≈ Depth (in) Beams seemed too deep. Checked reactions at ends of beam with hand calculations. Double counted self weight in computer design.

20. Foundation underdesigned. Forgot self weight of structure. Remembered later.

21. Common Problem: Structural layout does not match architectural and/or mechanical layout. Changes in layout not communicated by architect & not noticed by structural engineer. Will BIM fix this?

22. Overview • Categories of Errors • Experience of a Generation • Shared Experiences • Tools for Finding Errors • Teaching Tools for Finding Errors

23. Categories of Tools Hand Calculation Computer Analysis L 24 1. Comparisons (23 of 87) 2. Rules of Thumb (7 of 87)

24. 3. Visualization (5 of 87) 4. Other (14 of 87) • Procedures • Reflection

25. Similar Projects Oops, holes don’t line up! 5. Previous Experience (22 of 87) 6. Field (14 of 87)

26. 10 Quick Checks • Is the deflected shape consistent with what was expected? • Are the moment diagrams consistent with what was expected? = Identifying Features

27. 10 Quick Checks • Does the building weigh what you anticipate? • Does total base shear equal total applied lateral load? = Checking Equilibrium

28. 10 Quick Checks • Do beams deflect more than permitted? • If most beams are the same size, why are others not? • Is the beam depth consistent with standard rules-of-thumb?

29. 10 Quick Checks • Do the connections and bracing provide a continuous load path? • Do connection details match the assumptions used in the analysis? • Are the primary structural member sizes similar in similar projects?

30. Overview • Categories of Errors • Experience of a Generation • Shared Experiences • Tools for Finding Errors • Teaching Tools for Finding Errors

31. Teaching the New Generation • Many firms do so informally • Not in textbooks • Integrating at undergrad level

32. Learning the Tools Can be Taught • Comparisons • Rules of Thumb • Visualization • Other • Previous Experience • Field Procedures Over Time Least Preferred

33. Integrating Into Classroom • Emphasize approximations • Equilibrium always satisfied • Features of behavior

34. Approximations Situation: The roof shown experiences snow load with drifting adjacent to the AC unit. The resulting distributed load on member AB is shown. 300 plf 100 plf 100 plf A B 50 plf AC Unit A B 10 ft 5 ft 10 ft 5 ft Plan View Objective: Find, approximately, the peak moment and shear experienced by member AB.

35. Equilibrium

36. Features of Behavior Situation: A simply supported beam with a cantilevered end experiences uniform distributed load. Objective: Construct the moment diagram. Identify at least four features of the diagram that suggest you have a reasonable solution.

37. Impact In Classroom Structural Analysis I Structural Analysis II

38. Summary • Some errors not tolerated • Tools for finding them • Tools passed on informally • Tools can be taught Center for Structural Engineering Education www.rose-hulman.edu/csee

39. Acknowledgements Sponsor Grant: DUE-0341212 Participating Firms Questions or comments: james.hanson@rose-hulman.edu