A rant against the excessive use of FEM and FEA

A rant against the abuse of FEM and FEA

A real-life case, happened today:
A 120+ m long ferry, currently in design stage. There is an aperture on the side of the superstructure (10+ meters above the water) which measures 1000x2000 mm. We have communicated that the height of the aperture should be increased by 400 mm, bringing it to 2400 mm, in order to accommodate a telescopic crane for shore connection and stores.

The reply of the design office of a renown shipbuilding yard (cannot reveal it's name):
- "it can be done but would require redoing the FE analysis of the ENTIRE SHIP, and it has a price."



I remained stuck for a few seconds, trying to comprehend what was he talking about. The whole lateral area of the ship (just to give a reference number) is more than 3000 sq. meters, and the existing aperture is 2 sq. meters, which makes 0.07% of the total. The required increase of the aperture equals to 0.01% of the total lateral area of the ship... How the hell can it require a rebuilding and redoing the FEM of the whole ship?!?
So I replied:
- "Look, it is just an additional aperture of 0.4 sq.meters in an area of the superstructure which already has an aperture of 2.0 sq. meters and which is not interested by any structural reinforcements, just plating. Do you seriously believe this modification might compromise the structural integrity of the entire 130-meters ship? Well, perhaps it could actually - if the ship is under pressure like a balloon."
- The voice on the other side of the line: "I understand you but the classification society wants to see the FEA redone for each modification of the ship structure."

I have tried few more times to convince him that it is a nonsense. Just a decade or so ago a similar issue would have been dealt with directly in the yard, with a gas cutting torch. But my arguments were useless, their position on this was as rigid as a frozen dung.

At the end of the story, I will have to invent an ad-hoc folding, telescopic, unnecessarily-expensive crane which will fit the existing aperture, wasting a big number of design hours and client's money - just because those folks use and abuse the FEA for things which shouldn't require a FEA at all. I am still wondering if it is true that the class society requires them to redo the FEA for such a modification, or was it just another way of passing the buck to us because they didn't want to change the drawings...

End of the rant. The next step will be a referendum against the abuse of FE analysis in design offices.

Cheers

 

I would go directly to the officials and ask them. They might say don't worry about it. The worst they could do is confirm the design office position.

Or enlarge the hole and also add localized reinforcement to compensate so it would be so obviously stronger than it was originally that only a certified lunatic would object and require the whole ship to be re-analyzed.

 

The point is that rational arguements have not gotten anywhere. Another one is not likely to work with the attitude displayed.

I have had great success with going to the official source for other subjects. Generally the irrational statement is proved to be irrational.
Of course you risk jepordizing your working relationship with the boat yard.

The last of course is that redoing the FEA does not actually mean the entire FEA, just the local area. But what has to be done to prove the change to the classification society?

Too late now I suppose.

 

That is the point. Our client is actually the shipyard.
Hence, even if I might prove them technically absolutely wrong (the size of the required modification is probably smaller than the mean size of the mesh elements used for the FEA, just to say one), contacting the CS officials might be a counter-productive move. So at the end it will probably be a new crane and an extra cost for us, them and their client.
Tomorrow will try another round of negotiations...

 

All of the major costs to make the model and analyze it have already been spent. Making a simple modification to the existing opening and re-running the analysis is small potatoes at best. The SY sounds like it has already exceeded its design budget and is resisting doing additional work.

 

A ten minute conversation that might save thousands of dollars is not an option because of ....egos ?

I don't understand, if the client is the shipyard, and re-analyzing the ship will cost the clients money, or re-designing the crane will cost the clients money.....


Oh well, at least it doesn't come out of your pocket, I hope.

 

Some of the money will probably come out of our pockets too. We should have been more careful and notice this before the commercial offer was made and the contract was signed with the yard. They are a good client, and we will find a solution. Fortunately, this particular crane is just a minor part of the supply for this ship.

The point of the OP was their claim that a re-calculation of the FEM of the whole ship was necessary because of that microscopic (relative to the size of the ship) modification. It made us have a good laugh here in the office, when the story of the phone call spread around.

I guess the moral of the story is - too much FEM can hurt you.

 

You'd better hope the seagulls defecate evenly over the ship. Too much seagull poo in one area could stress the structure to breaking point.

 

Yes, I hope that situation was properly modeled in the FEA of the ship...

 

It sounds like this isn't a case of excessive use of FEA (a full ship FEM for a large, probably high-speed vessel makes plenty of sense), but rather a suboptimal use of engineering judgement when it comes to this design modification. It's easy to slip into extreme rigidity when you're in a later design phase, after you've finished analyses like these; this is certainly a better attitude than the opposite.

Are those your actual emails? If I was in your shoes, I might have tried to pull out the good ol' Dale Carnegie tricks earlier on to maximize my chance to win someone over to my side. I know what it's like to be furious when someone someone threw a wrench in your plans and that it's hard not to ridicule that which is ridiculous, but in this case I think you would have had a better chance of doing what's best for you, your customer, and the ship if you had taken a more understanding tone right off the bat, even if someone wasn't using their brain like they should be.

 

We don't know anything about what has been done, so I can't agree. Depending on the methodology and software used, re-doing an analysis with even small changes can be extremely laborious. Even if re-running the analysis is trivial, if a finite element report needed to be updated with dozens or hundreds of screenshots and a lot of numbers changed, that is a very expensive, frustrating process.

As OP noticed, re-doing the FEA is almost certainly not appropriate for this case, so this is not actually an issue, but I disagree with the assumption that the potatoes are small.

 

Simple.

You ask for the letter/email where Class asked for such. If they baulk and say it was just a telephone call, or on-site surveyor's comments, then that is a clear indication it is not necessarily true. So, the follow up Q is ..ok, please quote the actual Pt.Ch.Sec of the rules stating as such which must have been referenced by the surveyor!


See...simple

 

This happened so often in automotive we had unwritten procedures for coping. There was always talk about the probability some other department or supplier would trip a recalculation we could slip in on.

Somewhere behind it all is management or finance putting their foot down that the design absolutely must match the engineering analysis supporting it -not some previous revision, and a pack of lawyers poised to skin them alive if they can find a deviation. And deep in the bowels of the company there is an engineer responsible for that FEA who can make the result anything you like with some creative mesh work.

 

Oh, there was another phone conversation with them, soon after, done by another senior engineer. I think the good ol' Dale Carnegie has probably turned in his grave several times in few minutes during that one...

You are probably right. More diplomacy was likely necessary, though their stance on that issue was very rigid right from the beginning. At least we have had our 5 minutes of sincerely cynical comments and observations, both us and them. And it has ended with a friendly attitude, as it always does. Like I said, we will sort it out. It is partly our fault, so I cannot complain too much - apart regretting the good old days when these things were resolved on board in few minutes time by a couple of good steel workers...

 

Daiquiri, Is there any way to transfer the burden of solving the entire problem, not just the doorway, to the yard? (such as a different cast of characters from the previous go-round) Maybe cheaper than fixing a door or reworking the gizmo. This is an unnecessary interface problem. There is no apparent reason for two separate entities to to have to coordinate on a basic piece of kit. Given a choice between driving the wagon or carrying the load themselves, most will drive the wagon.

Of coarse there is one way - you just send them the crane that doesn't fit