Review of

S.H.S. Albracht, E. Blommert, Shell structures practical, Final report, course CT3280 Shell Roofs, Delft University of Technology, Department of Civil Engineering and Geosciences, 2024, online: phoogenboom.nl/B&B_schaal_report_1.pdf

by dr.ir. P.C.J. Hoogenboom, Delft, 17 February 2024

 1. Handed in almost on time (2 Feb. 2024). Preface is to the point. Excellent summary. We used to have no figures in summaries, but conventions change. You could have ended the summary with a sentence that your design can be safely build with a span of 70 m (or another value). This would be interesting to readers, wouldn't it?

 2. Clear text structure. Complete report. Excellent references.

 3. Page 5. This photo is taken from the IASS journal. You need to refer to this source. Every figure that you did not make yourself needs a reference. Otherwise, it is plagiarism, which can really get you into trouble. Famous example: prof. Diekstra, Leiden University. The other figures in the report have proper references.

 4. Introduction. Write in the past tense because for the readers the work is done.

 5. Very good that two clear objectives were formulated. You could have formulated an overarching objective: to figure out how to design by model testing. (You might find it strange that the teacher did not just tell you this objective clearly. But this is how live works; you do things; people around you do things; smart people put in words what we apparently aim for; then others say, "I know this."; on a deep level they do; but they could not formulate it; until you did.)

 6. The content of the chapters is explained on page 2, page 4 and page 5. A bit much, isn't it?

 7. Chapter 1. Clear reporting, Very good: your team immediately started trying out things. Was the paper mache necessary? Perhaps, concrete could have been casted directly onto the yoga ball.

 8. The design is not very challenging. Safe choices were made. On the other hand, this team was fast and did not copy anything from other teams.

 9. Section 2.2. There is no reason to think that the wire mesh yield stress is low. It is just steel. We could have done a tensile test on the mesh. The joints between the mesh parts should have overlapped. This is called development length (verankeringslengte). The force is transferred from one bar to another bar by development length. Nonetheless, since there was no development length, it makes sense to ignore the effect of the reinforcement on the shell wall tensile strength.

10. Figure 2.3.2. is the same as Figure 1.2.9. We do not do this in technical reports. The text can just refer back to a previous figure.

11. Comprehensive description of the shell construction process

12. Complete description of the test setup

13. Complete description of the material tests

14. Page 21. Two things go wrong in the calculation of the modulus of elasticity. 1) fcm Is the mean concrete strength as determined in the material tests, fcm = 8.1 N/mm2. 2) The characteristic strength fck is obtained by subtracting 5 not by adding 8.

15. Good hand calculation of the buckling load, correctly including the knockdown factor.

16. Page 22. In the top of the dome is compression, as Figure 4.1.2. shows. So, the dome top will not fracture.

17. Page 23. The deflection of a distribution wood is checked. Why not its stress?

18. Page 24. Good overview. Very good hand calculations, despite the above exceptions. Everything is checkable, which is very important.

19. A span of 35 m is 51 x 0.688 m. So, the real dome is 51 times as large as the model ... nice. 

20. Good idea to have small pictures in the chapter and large ones in the appendix. Had not seen this before.

21. Page 25. The Von Mises stress is accurate for failure of metals. It is better to look at the principal stresses.

22. What element size was used in the analysis? Was it small enough?

23. What imperfection was used in the nonlinear analysis? Others must be able to reproduce this analysis.

24. Page 27. Good that a comparison between the hand calculations and the computation is made.

25. Very good photos of the damage to the ropes. Solid analysis of what went wrong.

26. Why not a load-displacement diagram? Why not compare this to the computed displacement? Would have been an interesting check.

27. Chapter 7 After testing, starts with a summery. Was the first summery not good? There is no need for two summaries.  

28. The conclusions should refer to the objectives in the introduction.

29. The arguments, on which the conclusions are based, are clear and valid.

30. There could have been recommendations on how to build the real shell. How to make the formwork? How to make the right shell thickness? 

31. Complete references