Brief History of Computers in Structural Engineering

1976- This is the first time I’m working with computers. This is my second year at Tehran Polytechnic (Amirkabir University of Technology), and as students of "Numerical Analysis" we have access to an IBM 370 mainframe. The input data is given in the complex formats of Fortran IV using the punched cards, and the output device is a large size printer. The system as a whole occupies some 4000 square feet.

1979- We use the same computer, now in analysis of our steel and reinforced concrete final projects, using the only reliable software of this time, STRESS II. This program is written in FORTRAN IV using the stiffness matrix method, for plane frames only. It takes some one hour to complete the analysis of a 5 storey 5 bay plane frame, for 2 load cases and 3 load combinations, enough for elastic (allowable stress) design.

1983- The first job of my career is started in an office, fortunately equipped with the first personal computer I have ever seen and used: Radio Shack 16, with 16 KB memory, and a BASIC86 interpreter. People are strongly advised not to say kilo bytes (to computer experts, it's a crime!) as kilo means 1000, but this new K means 1024 instead. My supervisor, the chief structural engineer, has prepared a few useful programs for analysis of plane frames using Takabeya Method (a moment distribution method just like Cross and Kani, but much easier and faster), and design of steel and reinforced concrete members using AISC and ACI respectively. There is not any local code in Iran yet, so most engineers use the American codes (UBC, ACI, and AISC), while others use British (BS), French, or German (DIN) codes of practice. A few months later... an extra memory of 48 KB is purchased and added to the system, which makes it 64 KB in total. Wow... 64 K-bytes! Who could even think of it?! Now it is time to start writing a brand new program for frame analysis using stiffness matrix method. It works great for 2 to 5 storey buildings. But in solving the equations, we'd rather use Cholesky Method which is faster than Gauss Elimination Method. The results of the analysis are exported to a cassette, and then imported to design programs. I’m so happy that with these programs, we are not using the computer as a typewriter!

1984- Same firm: a big step forward, with an NCR DM5 personal computer, using an MS Dos operating system for the first time, with 640 KB memory now called RAM, with a couple of 5-1/4" floppy disk drives, and 2 hard disks each 5 MB. It’s unbelievable... 10 MB of storage! It’s enough for a large size m storey n bay frame... and even more unbelievable is the speed of access to the hard drive, you say what file you need, and there it is; you don’t have to find the beginning of the file as you did on cassettes before, and you don’t have to verify the file loaded either. It’s now possible to store a few of the photos which would be taken some 20 years later by digital cameras, in the hard drives! The processor is now called 8088, and you should still wait a few more years to hear about the 80X86 series of computers, known as IBM compatible.

1991- The new company I join, has recently purchased the first 80286 generation computer, still 640 KB RAM, but with a hard drive of 40 MB. MS Dos 5 is released, and the very first version of AutoCAD is under curious investigations by some companies. There is not any structural analysis reliable software in the market yet.

1993- The processor is upgraded to 80386, and the hard drive is increased to a storage capacity of up to 80 MB. The R10 AutoCAD is now widely used in engineering offices. PC Tools and Norton Commander are now replaced by a new software released by Microsoft, Windows 3.1. It does whatever needed in managing the files, that is, copies, deletes, and renames files very easily at a glance. Fortran, Cobol, and even Pascal are almost forgotten, and everybody talks about new versions of BASIC which are called Quick Basic and GW Basic. First versions of Berkley SAP80 structural analysis software is capturing the market. But the format of data input as well as output process is a real pain. It’s neither as interactive as STRESS II before, and nor as easy as the graphic versions to come after 2000, but is a powerful program in analysis of plane and 3-D frames instead.

1995- 90's versions of SAP are replacing SAP80. AutoCAD R12 is used widely. Not many rulers, pens, and erasers may be found in engineering offices any more! New Windows 95 is not only a powerful file managing tool now, but also an incredible operating system based on MS Dos 6.2.

1998- Processors are upgraded to 80486, and in a few months will be replaced by 80586 or Pentiums. AutoCAD12 is used worldwide, and Windows 98 is taking the place of the previous 95 version.

2001- High frequency processors and large size RAM's, along with GB hard disks are available now. Various versions of Windows, 2000, Millennium, and others are being tested everywhere. Most programs are now Windows oriented, and Dos programs are being forgotten little by little. After a short period, the first Windows based AutoCAD14 is replaced by AutoCAD2000. SAP2000, ETABS2000, SAFE, and many more analysis and design softwares are now used by structural engineers all around the world. Good bye FORTRAN formatting of input and output data! Try drawing your structure in a whole graphic environment instead now. Many structural member types such as plates and shells are now added to the analysis programs.

2006- Windows XP in two Home Edition and Professional categories are beeing used for years. Giga bytes sounds like jokes or kid toys now: 2 GB RAM, and 250 GB hard drive. Multimedia, DVD, flat monitors, it’s like a dream.

I often think how long it took to travel by jumbo jets instead of carts and horses, and how many years and months and days are passed since I was that much happy with my NCR DM5 with 640 KB RAM and 10 MB hard, which I used so efficiently with my own programs. It must be a dream. It is a dream... or is it?