Dr. Subramanian Narayanan- Award Winning Consultant, Author & Researcher
When I started my consulting company in Chennai in 1982, after my return from the then West Germany as Alexander von Humboldt Fellow, computers were just being introduced in India. At that time the cost of hardware was much higher than the cost of software. Moreover, no special hardware was available on these machines. We bought a non-standard Uptron Computer with 64 kB memory with 8 inch floppy disks (5.5 inch and 3.5 inch floppies were introduced later; By 2006, computers were rarely manufactured with installed floppy disk drives). I wrote analysis and design software to run on this machine. The quality of hardware and software on computers slowly improved with the introduction of Personal Computers marketed by IBM (though first introduced in 1981, it appeared in India around 1985).
Interestingly, we used only slide rule during our college days, which was difficult to use especially when very big or very small numbers were encountered. It is also noteworthy to remember that the famous Empire States Building, the 102-story Art Deco skyscraper in Midtown Manhattan, New York City, was designed by Shreve, Lamb & Harmon and completed in 1931 (see Fig. 1), without using any sophisticated tools or computers, but is still standing without any deterioration and with all its glory.
Compared to yesteryears, the tools available to the structural engineers have grown exponentially, both in number and in sophistication. They are also available freely and at a reasonable cost. Of course in the present competitive environment, the efficient use of these sophisticated tools is absolutely essential. It may be impossible to provide efficient and timely service if one uses the tools of yesteryears, such as slide rules and drafting tables. However, it is important to realize that the structural engineers of today might have lost some important skills along the way.
When using the early tools, the engineers had an understanding of the engineering principles and also had a feel of how exactly designs were really obtained. Thus, they were able to realize whether the calculations that were carried out were proper and also the solutions achieved were reasonable.
Now with the use of black-box computer programs, fast computers, BIM, and other tools, the structural engineers rely much on the computer outputs. Many engineers do not realize that precision does not guarantee accuracy. When we start our calculations with code defined live loads (which may or may not reflect the actual loads we may encounter in the structure), the dead loads (that are determined to the best our knowledge), and various assumptions made in the analysis and design, computations to the 8th decimal place may not be necessary (except maybe in the case of shell structures). It is necessary to understand that computers and software do not consider these facts and faithfully do the computations based on the algorithm with which the software is written.
Another problem is that the availability of software and computers gives the engineer the false assurance that anything can be analyzed and designed by them. Complex structures should be designed only by experienced engineers and novices should attempt to design such structures only after getting enough training or with the guidance of experienced engineers. For example, tall buildings, shell structures or space frames have their peculiar problems and should be analyzed and designed only under the guidance of experienced engineers. One should note that ‘pushover analysis’ itself is approximate and the results obtained should be considered carefully with a pinch of salt (Moghaddam and Hajirasouliha, 2006).
I will explain the situation with an example. Recently an engineer was asked to design a cold-formed parallel chord truss with cold-formed C-sections as shown in Fig.2. The ends of these trusses were connected to concrete columns using steel angles (see Fig. 3). The trusses had a span of 3.4 m, and loaded with a uniformly distributed load of 5.5 kN/m.
The engineer analyzed the truss using standard computer software, with 2 pinned end supports at the top chord and 2 pinned end supports at the bottom chord and got the horizontal reaction of about 18 kN and maximum deflection of 5 mm. Though the members were safe, he could not provide a connection for the horizontal reaction, as the capacity of screws was exceeded. He had provided slotted holes for the truss in the horizontal direction as he obtained a horizontal deflection of 2 mm. What is the solution to this problem?
I suggested him to remove the top pin support, as the bolts he has shown are only to keep the truss in position. I also asked him to provide only roller support at both ends, as the connection can move in the slotted holes. The problem was solved as there were no horizontal reactions. This shows that though computer software may be available, judicious thinking is required to arrive at a practical solution.
Advantage of Experience
In many cases, an experienced engineer will be in a position to give an approximate solution even to certain complex problems. Inexperienced engineers need to study ‘thumb rules’ in order to find out whether the computer results are correct. The old saying ‘garbage in-garbage out’ is valid for any computer output. The accuracy and correctness of any computer output are based on the assumptions used in the software and the input values. The computer will not recognize any wrong input and give output for the given data. Experienced engineers could easily check the computer results by doing some rough back-of-the-envelope calculations, or sometimes even looking at the deflection profile of the structure or by checking the reactions at supports.
More such thumb rules and practical tips for concrete elements may be found in Appendix D of Subramanian, 2013.
Another issue to be borne in mind is constructability and practical difficulties. For example, only steel bars of selected diameters are available in practice. In the same way, only selected sizes of steel sections are rolled and easily available in the market. Experienced engineers can easily realize these problems. Use of optimization techniques will result in some quantities for sizes, diameters, etc., which could not be adopted in practice due to the above mentioned reasons. In addition, whatever is obtained through computer analysis and design may not be feasible due to constructability problems or may result in congestion of reinforcement. Experienced engineers will be well versed with correct and proper detailing procedures than fresh engineers.
Finally, not all problems require sophisticated analysis and design. Almost any problem could be refined, for example by adopting a finer mesh in FEM analysis, indefinitely without providing additional value. Hence the engineer should recognize such situations and avoid making such refinement in order to save money and time. Moreover, simple structures are always better than complicated ones, as their behavior could be easily understood even during earthquakes.
The technology available to present-day structural engineers is exiting and more advanced. These tools are important, necessary, and essential. However, it is important to realize that experience will teach us not to entirely dependent on these advanced tools and also recognize the pitfalls of using such technology.
Moghaddam, H. and I. Hajirasouliha, "An investigation on the accuracy of pushover analysis for estimating the seismic deformation of braced steel frames", Journal of Constructional Steel Research, Vol. 62, No. 4, Apr.2006, pp.343-351.
Subramanian, N., Design of Reinforced Concrete Structures, Oxford University Press, New Delhi, 2013, 856 pp.
White, B.L., “Technology is not a Substitute for Experience” Editorial, Structure Magazine, ASCE, Oct. 2019, pp.7