Method to solve as-built, discrepant, steel structure issue 2

[MotoGP]

Greetings to all Structural Engineers:

I am a MechE and, only by the grace of God obtained my PE. As such, I can only consider myself very fortunate at having a venue, such as this one, to ask what all of you will find to be a much too easy problem. I beg both your attention and indulgence.

Scenario:
1) Given: a "section" of a mild steel structure with welded joints subject to dynamic loading (vis, a 145K gtw ship's hull).

2) The section in 1)(above) contains two (2) flat, rectangular members that are attached to each other at right angles. The approximate width and length of both structural elements are 20mm and 1500mm, respectively. They are both 300mm in height.

3) The subject structure is utilized to provide "stiffening" for the hull plating, and serves no other purpose, and is located within the confines of a cofferdam immediately forward of the main engine room.

4)One of the steel stiffeners has a 3:1 (length-to-height) taper on one side and is fillet welded (all-around) to the other stiffener that is oriented at right angles to it. The taper begins at 200 mm before the joint with the other sitffener

5) Field inspection reveals that in lieu of the 3:1 taper, it has instead been cut with a taper of approx 3:1.5. 3:1 is the design requirement.

6) One of our group's structural engineers looks at this scenario in-place, and asks my opinion. I state that the discrepant end should be cropped and re-newed using the same material, but with the correct bevel. To wit, the much older structural guys has a good belly-laugh, makes some inflammatory remark about my comparative youth and lack of structural knowlege, and then tells the shipyard QC guy that "it'll be just fine the way it is".

7) Later, I ask said structural engineer about how he was able to arrive at his conclusion so rapidly without any calculations, etc.(I do this while visions of those mystifying colorful, structural FEA diagrams flitter around in my head).

8) The response is along the lines of that his long experience is the "mother load" of knowlege behind all decisions that he makes, and also that he has not met anyone who can make these decisions as fast as he can, etc. etc.

9) Is this colleague a "loose cannon"? If not, then I have been straining at ways to figure out similar problems where not as much sturcture is utilized in construction as the design calls for...is there a resource for those in my position, or, is it truly only a lenghty experience that can help me to solve these problems? There should, I reason, be some point where up-to-"x" mm material "short" may work in a structure, but, beyond that point, material will have to be cut and renewed...right?...maybe?

It is now my only hope to have at least described this problem in a clear manner.
Very Best Regards,
(Name witheld due to embarassment)

 

[SlideRuleEra]

I'm afraid that I don't fully understand the technical problem that you have described (probably my unfamiliarity with ship constuction). However will offer some opinions on you colleague's behavior:

Sounds like is a little too boastful and would be more helpful if he offered to share his insight into his solution. However there is a nugget of truth at his claims.

From my contact with younger engineers, there appears to be a subtle difference, over time, in the way engineering is taught in college. In the 1960's and earlier (no electronic caluclators or routine access to computers) students were taught to look at complex problems with the thought process being:

What is going on here? How can I "transform" this situation into a simpler approximation that I can solve (typically with a slide rule)?

The result was often an approach that lead to "bounding" a problem, i.e. - I know that my first appoximation will "work" and that a second approximation will "fail" - Now how does the real problem "fit" between these two approximations. Perhaps your collegue "recognized" that even the worst case geometry of your problem was adequate for the forces involved.

An engineering education these days seems to focus on how to use the modern tools (usually software) now available to correctly address a complex problem. This results in much better solutions, but does tend to minimize the necessity to understand "What is going on?"

 

[MotoGP]

Slide Rule ERA:

Thank you for your constructive opinion. I'll do my best to sober up and see the "big picture"...quite hoenstly I am not sure where to begin, rather, how do I go about "prioritizing" what is critical to adhere to in a steel strucutal design vs. those things that that are of lesser importance. At my stage of the game, it seems like EVERYTHING is important. I try to make sure that the material grade and thickness(es) of the pieces used are i.a.w. the approved design, and that the correct welding practice is followed. Next, I look for alignment problems, and areas that my require fairing or line-heating to "clean-things-up" a bit. If any penetrations are made in the steel I then check to see if they are in the correct location and cut to the appropriate dimensions. Of course I check to see if all the pieces are installed and assembled correctly. Finally, I then look at the end connections, weld quality, grinding, etc....look for unauthorized repair work and the like.

Anpther example that I am having the greatest challenge with is looking at a bracket end connection, let's say, with a flanged longitudinal stiffener. the connection should be, i.a.w. the design and shipyard standards, 35mm below the flange. No tolerances are given, and I can't seem to find anyone who is interested in committing themselves to an "un-official" set of tolerances. Yet, I find in cases anywhere from 40mm up to 50mm cut away. Is this critical? is 40mm accpetable? Is 50mm requiring repair or it is also accpetable? The structural design people say that I must follow the dimensions on their drawings, but, when things get fabricated, there then seems to be invoked a "black art", by a comparatively more knowledable structural engineer, who, somehow, mystically, decides whether the as-built condition is accpetable or not. I want to undersand the "black art"...but am told that this is matter of "experience"...what experience? I feel that there should be some method of solving for these problems to decide what is acceptable, but I have not yet come across anything.
Anyway, thanks for taking the time to help me out.

 

[SlideRuleEra]

MotoGP - Structural steel design is a good example. I would suggest moving investigation of member connections to the top of your list. The vast majority of failures can be traced to a problem with a connection - members can fail, but rarely catastrophically. Then check for yielding, lateral-torsional buckling, shear and deflection. If you have high point loads, look at both web & flange local buckling.

As for "what" to check for an "out of spec" situation, try to step back from the details and make some assumptions such as:
1. Everything has "understandable" supports (simple, fixed, cantilever) - choose the one that looks most reasonable for each situation.
2. Make "understandable" assumptions on loading (uniformly distributed, point, load increasing uniformly) - again choose the one that looks the best.
3. Perhaps you can assume that curved members are actually straight.
4. For members with multiple supports, take a look at a series of members, all with simple supports.

These are just hypothetical examples. A good place to get ideas is from the "Beam Diagrams and Formulas" in the AISC Manual of Steel Constrution.

Now for the "tricky" part - try to find a qualified mentor who is willing to take the time to show your what is both "right" & "wrong" with your approach to a problem. Also read about what is happening in your industry - studying the details of failures is usual more educational.

Be Patient. I don't know how to speed up the process, but slowly you will start to realize that things "make more sense" than they did before.