Addressing Source Credibility and Clarifying Guidelines for Engineering

Just noticed (as a follow up of a recently terminated one) that many topics posted in 'Engineering' actually does not conform with the general rule about respectable sources, causing confusion.
Sure, engineering has a way to say 'see? Works.' and it's totally (!) respected in the actual attitude but somehow I feel that a clarification about the acceptable starting points (respectable journals or an actual engineering feat*? ) might be useful.

Opinions?

* well, sometimes: blunder

CC @berkeman @russ_watters @jrmichler

Rive said:

Just noticed (as a follow up of a recently terminated one) that many topics posted in 'Engineering' actually does not conform with the general rule about respectable sources, causing confusion.

Can you post some links to example threads? Often in Engineering threads, the subject is not so much controversial, and more of a question about how to do calculations, etc.

berkeman said:

Often in Engineering threads, the subject is not so much controversial, and more of a question about how to do calculations, etc.

This.

Much of engineering theory is from the 1800's. The theory of electric motors was well understood by the early 1900's, as was the theory of heat transfer. The theory of fluid dynamics was well understood by 1904 when Prandtl started the theory of boundary layers, then pretty much fully developed by 1955 when Schlichting published his Boundary Layer Theory. The Moody diagram is from 1944. The theory of semiconductors, along with the theory of semiconductor analog circuit design, was well understood by 1964, when General Electric published their Transistor Manual.

An undergraduate curriculum in mechanical, electrical, chemical, civil, or aeronautical engineering covers principles that were effectively fully understood 100 years ago. These principles are well covered in undergraduate textbooks. Those textbooks differ in their quality of communication, but not in their coverage of basic theory. Much of this is still true at the graduate level. Engineering problems for which journal references are appropriate are rare.

Journal papers tend to be so narrowly focused that they are difficult to understand and apply. When I was working on active vibration control, I read many journal papers, and found almost nothing that was useful. So I ended up creating a basis state space controller, then developing a hand wavy technique to tune it and find the most stable operating point. That operating point was not the point of optimal control. That technique is partly documented in US Patent 5,983,168.

I guess it may be worth clarifying that engineering isn't science, so you don't need peer reviewed sources but you do need...something real? I'm not sure how to define what an acceptable source should be.

russ_watters said:

I'm not sure how to define what an acceptable source should be.

The proof is in the pudding. Answers to engineering questions are more like education than research. No one in the physics forums cites papers when asked to explain using a Lagrangian to solve a pendulum. Same with calculus questions. There are better educational sources, like text books, than original research.

In my experience here, even in theoretical physics, peer reviewed papers aren't actually used much to explain things. I mostly see them as a way to challenge questionable assertions, or as genuine questions about a particular paper. How often is an Everett paper actually used when someone asks about MWI? Einstein for GR? Nope, maybe Misner Thorne and Wheeler. I'm not convinced, in practice, that "acceptable sources" are actually used that much in any of the forums.

Engineers have text books too. In addition, back in my day, there were some excellent publications from device manufacturers and conference proceedings. The whole academic publishing world isn't as relevant for people that have to actually design products to sell. We aren't paid to publish, as a rule. As @jrmichler said, most true research papers (i.e. peer reviewed, high impact journal) aren't really of much use to us. You're more likely to get shown how to solve a problem, or given a counter example in our world. A clear explanation of something hundreds of engineers have already done is more productive.

As a scientist, I consider some ideas and concepts to be really confirmed (as opposed to being not falsified) when they are repeatedly used to successfully engineer things up.
Each production is like a little experiment testing the underlying concept.
Failures standout in engineering and engineering is not just limited to highly controlled situations.

Thanks for the opinions! And sorry for the delay.

berkeman said:

Can you post some links to example threads?

This was the one which planted the bug: link

The cited paper is just not fit for scientific discussion (here), that much is clear: even with plenty of stuff from the paper is actually available from different (usually engineering, so: not peer-reviewed&scientific - science apparently just passed the topic already in the pre-internet era...) sources.
Really, kind of the textbook feeling. No sources, because things are in the textbook phase already.

But there was this topic linked there, with starting point even less conform with the requirement of 'respectable source', but still open (just as like most of the engineering section).

Having both topic on the table in the same time what makes this kind of uncomfortable. Make things look like just as an excuse. And I'm not really comfortable with excuses.

Rive said:

This was the one which planted the bug: link

One of the big problems with that thread start was that the OP was claiming a 50% efficiency gain over existing high-power hydropower installations, which is clearly not possible (given the >90% efficiency of such installations currently). There were other issues, but that is one of the main reasons for asking for valid references. If the OP wants to make extraordinary claims, they will need to back them up with a reputable reference.

The OP then later moved the goalposts a bit and said effectively "Oh, I meant low-flow hydropower installations", which was pretty different from their thread start. In the end, the thread was doomed by several issues, including copyright violations in the OP's posts.

Rive said:

But there was this topic linked there, with starting point even less conform with the requirement of 'respectable source', but still open (just as like most of the engineering section).

That thread was more about discussing the interesting concept of tapping an existing hydropower source that was not utilized yet, and obviously had to discuss the fact that extracting hydropower from those existing irrigation canals had effects on the flowrate. I don't think there were any outlandish claims being made by anybody in the links in that thread, and they all seemed to acknowledge the complications and issues.

The issue I finding slightly disturbing is not with the result, but that the reasoning for keeping the second topic (and plenty of similar around the Engineering) open is not reflected in the Terms & Rules: even with the practice already pretty cleanly admitting and supporting them.

Do you have a suggestion for a modification of the rules to make this work better? Constructive suggestions are always welcome by me.