Review of

Y.J.C. Bot, F.C. Cauwels, J.L.J. Wieringa, Shell structure design, Group 2, Report for course CT3280 Shell Roofs, Delft University of Technology, Department of Civil Engineering and Geosciences, 2024, online: phoogenboom.nl/B&B_schaal_report_2.pdf

by dr.ir. P.C.J. Hoogenboom, Delft, 19 February 2024

 1. The reports structure is clear and in the right order. In most of the report the language is okay. There is no need for the date on every page. Do not put units between brackets []. Handed in on time (1 Feb. 2024).

 2. The Summary is complete, to the point and interesting. However, it does not read well at all. The first sentence contains several language mistakes, which makes a bad first impression. The second sentence states the objective of the work "... design a shell structure ...". This could have been formulated clearer. Actually, almost every sentence of the Summary can be improved a lot. This is not an acceptable level of English for engineers. 

 3. Do you really need to copy the first sentences of the introduction from somebody? .. and there are language errors in it. You need to learn to write your own sentences.

 4. The research question in the introduction is " ... examine if the design of a real size roof shell will be strong enough ...". This is different from the objective in the summary. Be precise in formulating the objective or research question.

 5. Figure 7 and 8 are identical. Something went wrong with adding one of the figures. Did nobody check the report? It is just 16 pages.

 6. Page 4. The final design is really beautiful, I think ... and it is buildable too. Excellent.

 7. Page 6. Hand calculations: The thickness of 12 mm does not agree with what was built. The stress is not checked. The q load on a lath is not calculated. The total load on the shell is not calculated. Hand calculations were not given attention at all. There was plenty of time and assistance to do this. Would have been good to check the computations.

 8. I count 40 laths of about 1 m length. The line load q = 0.204 kN/m. So the total load is 40 x 1 x 0.204 = 8 kN. This is much too large. Is there something I misunderstand? Did you do simple checks by hand?

 9. Page 7. Smart to calculate the surface area like this. 

10. Page 8. How large is the imperfection amplitude? Results need to be reproducible by another engineer. If the imperfection amplitude is not in the report, he cannot check it.

11. You analysed the real size shell. Did you switch off self-weight before the analysis? 

12. Excellent computations. However, the reporting is too short. What is the largest stress at 1 kN/m2 ? What is the largest deflection at 1 kN/m2 ? What is the buckling load factor at 1 kN/m2 ? How did the first buckling mode look? Does this agree with hand calculations? Which was governing; stress, deflection or buckling?

13. Good description of the construction process. I wonder whether the hot water and drying was necessary. Good that this comes back in the discussion. Did you use 30 minutes drying glue or 4 hour drying glue?

14. The photos show many extra laths compared to the design. Why was this?

15. Good description of the load distribution system. Valid check of the rope strengths. You could have checked the stresses in the wood too.

16. Page 12. Clear and correct calculation of the collapse load prediction. (The hand calculation result is obviously ridiculous.) In retrospect, the 50% reduction was a mistake, wasn't it?

17. Page 12 and 13. Solid description of the failure. Spectacular youtube movie. Good job editing. Only 20 views at present. I will refer to it.

18. Interesting load-deflection diagram. Highly nonlinear behaviour

19. Solid no-nonsense discussion. This test was the first. We were learning.

20. Interesting and true conclusions

21. Literature ... Who is "I. Concrete"? :-) If the author's name is not known, start with "Anonymous". The rest of the references is fine.

22. Very informative appendix