Engineering without Computer Models/Simulations?

I was reading an article on experimental nuclear reactor concepts, and the author mentioned that the development of high power computers in the 1970s was greeted with enthusiasm by nuclear scientists, as they would be able to experiment with far more designs than if they had to build and test physical demonstration reactors. However, very few of the designs have been built, even ones with great promise.

While some of this may be due to decreased funding for basic research, especially nuclear energy research, similar trends can be see in other areas. For example, there were many experimental aircraft built in the United States through the 1960s and mid-1970s, and afterwards there haven't been as many. When things are build, they tend to be refinements of existing ideas (for which there is significant data), and not many new ideas. In contrast, the Soviet Union had less computer modelling and simulation capabilities relative to other developed countries and continued building physical prototypes and demonstrators for many concepts. While it may be correlated to a different approach to research, it's an interesting observation.

If computers were not as powerful, or if model and simulation theory wasn't as advanced, might there be more experimentation in certain fields, such as nuclear energy and aviation? After all, information on performance would only be available in general terms before building a physical test model.

I think there would be less progress, since modeling costs much less than building real life test beds. The reason engineers love modeling is because it saves time and money.

Computer modelling has enabled the project managers, accountants and politicians who want to guarantee optimum performance. Engineers know that computational modelling helps you get to 'highly likely to do the job well enough' much faster, but the rest of the organisation decides that it means the engineers can just work on 'certain to do the job perfectly' instead.

Do away with computational modelling, and the bean counters will have to accept - that sometimes things won't work quite as well as they'd hoped for. Project definitions are likely to stay away from the bleeding edge of technology to derisk them, rather than opting for death by computer model, so you'd probably see a 70% solution every 5 years rather than a 100% solution every 20 years.

This is expensive and slow, of course... but at least we'd have discovered lots of ways of doing the job that don't work, and built loads of cool prototypes along the way. Engineers might say that we love computer modelling whilst on the clock, but off the clock we'll gleefully admit to liking building stuff then trying to break it.

Less progress, more development cost, more production cost, etcetcetc.

Computers - the more powerful, the merrier - is big in engineering. Just some samples: with finite element method, we can predict the critical points of a structure, prior to any testing and redesign it "on the fly" if needed. CAD/CAM is a blessing - not only make the design process and production adaptation a hell a lot easier, but make it cost effective and opened up possibilities we could have dreamed about before. Networked logistical systems do wonders... visit Rotterdam container terminal, and its a miracle.

Not to speak about the documentation making

Of course, someone's eventually going to ask the inconvenient but apt question 'what's the PoD'.

So what is the PoD?

MattII said:

Of course, someone's eventually going to ask the inconvenient but apt question 'what's the PoD'.

So what is the PoD?

Click to expand...

How about one I've used before, with the government deciding not to fund the $10.4 billion (source) Semi-Automatic Ground Environment program (SAGE), which would be the equivalent of $87.82 billion today. It created a computerized air defense system for the United States, and IBM was the prime contractor. It helped make the company what it became and helped establish the modern computer industry.

Delta Force said:

How about one I've used before, with the government deciding not to fund the $10.4 billion (source) Semi-Automatic Ground Environment program (SAGE), which would be the equivalent of $87.82 billion today. It created a computerized air defense system for the United States, and IBM was the prime contractor. It helped make the company what it became and helped establish the modern computer industry.

Click to expand...

Still wouldn't stop what Sperry, Control Data, DEC, Burroughs and even IBM was doing on the commercial side, that would lead to Seymour Cray leaving Sperry for CDC.

When they wouldn't do the machine he wanted, he left to make his own company to make a 'Super Computer' in 1972, that ended up being sold to Los Alamos over Lawrence Livermore Labs attempts to buy the first one three years later.

Cray expected maybe a dozen companies would want a machine that powerful, but he sold 80 of them, costing up to 8 million USD each.

Even if you would have him hit by a bus, that demand was still there, and there would be a machine to fill it.

marathag said:

Still wouldn't stop what Sperry, Control Data, DEC, Burroughs and even IBM was doing on the commercial side, that would lead to Seymour Cray leaving Sperry for CDC.

When they wouldn't do the machine he wanted, he left to make his own company to make a 'Super Computer' in 1972, that ended up being sold to Los Alamos over Lawrence Livermore Labs attempts to buy the first one three years later.

Cray expected maybe a dozen companies would want a machine that powerful, but he sold 80 of them, costing up to 8 million USD each.

Even if you would have him hit by a bus, that demand was still there, and there would be a machine to fill it.

Click to expand...

I'm not looking to prevent the development of powerful computers, computer modeling, and computer simulation, but delay it by a few years. SAGE was a large hardware purchase, and it required so many programmers to write the code that it helped formalize their training.

The airline reservation system was one of the first offshoots of SAGE, and it also happened by a chance meeting of an airline executive and someone working on the project. So one of the first major corporate initiatives would be delayed too.