how much above "capacity" you feel comfortable going over? 2

[AskTooMuch]

I have an existing steel beam analyzed using LRFD for additional dead load. I used Risa3d and unity ratio is 1.2 for bending check. Say I already sharpened the pencil as far loading is concerned.

 

[AskTooMuch]

I can easily use 1 or even 0.9 UR all the time and sleep at night better.

But I have a responsibility IMO as an engineer to do my due diligence to not ask for modifications which definitely cost money just so I don't have to worry about lawsuit.

 

[ThomasH]

One very simple question. If you say "no, it is not possible", what is the consequence for the client?

Thomas

 

[Ron247]

just so I don't have to worry about lawsuit

There does not have to be some failure totally related to your design for you to become involved in a lawsuit and lose. You only have to have "contributed" to the problem and the due diligence you cited becomes a big factor in that outcome.

As I stated previously, can your Client control other loadings during the maintenance operation? If so, I would rather take that route than using a CSR of 1.2 unless I can cite a recognized standard that states it is acceptable.

For those out there who do not do much expert testimony or get involved in legal, follow this legal train of thought and ask an attorney if I am correct.

Two identical businesses are located side by side. Company A does a study to determine hazards present in their company and after assessing cost versus risk decide to not correct any problems. Company B chooses to not even do a study, they decide to just take whatever unknown risk is present. In other words, both companies have identical risks. When the same accident occurs in the 2 companies the legal outcome could be drastically different. Company A may lose their lawsuit but Company B will also lose theirs and very possibly have a much higher penalty. Knowing a risk and choosing to accept the risk is viewed less punishable than just not even bothering to see what risks exists. Company B did not perform Due Diligence whereas Company A did. Due Care is the 2nd piece. Due care is what did you do about your identified risks? Both chose the same Due Care path.

 

[Ron247]

This one is a long shot but sometimes works. Field measure the actual beam near the max load point. Sometime tolerances go in your favor. Especially older beams. Depth and flange dimensions are easy to field measure generally. Those 2 more affect your bending.

 

[BridgeSmith]

Some of the guys that have more familiarity with the legal aspects please correct me if I'm wrong, but it may come down to acceptable risk, right? Obviously, injuries or loss of life is not acceptable, so there can't be anyone under or around that area of the floor while the equipment is in place. However, if the risk to the structure is acceptable to the client/owner, then you can let it be their decision, so long as you have presented the risks to them. My guess is that, when presented with the risks, the client/owner will opt to utilize temporary shoring during the times when the equipment is in place.

 

[AskTooMuch]

Just for info, I'm one of the Maintenance Engineers of the plant. There is really no client because I'm part of the client. Operator of the unit asked if they can put the equipment on the floor during maintenance/turnaround. The equipment is currently on the same level but supported by a separate structure. The floor goes around the structure. They plan on lifting the equipment via tall crane and temporarily put on the floor. I'm still on preliminsry stage of the design but I will ask what kind of live load I'm expecting during turnaround. I assumed 100 psf initially.

I'm not really asking how to design the floor.
My question was how much above 1.0 you are "ok".
I think the link EIT gave already answered my question.

Thanks,

 

[Rabbit12]

I think the answer to your question is there is no hard fast rule. It depends on lots of factors; failure type, load frequency, and the engineers gut feel.

It's just too vague of question to get blanket answers. Personally if I sharpened the pencil to a needle point and it was still 20% over, I'm not sure I'd let that go. It'd be really tough to explain that in a grieving widow lawsuit.

 

[AskTooMuch]

I'm not ok wth 20% too, that was assuming the equipment load is at tbe center of the beam, as I move the equipment closer to column I can get less than 1.0 UR. Of course the shear on that connection must be checked as well. I think I already got what I need and this will be my last reply on this thread. Thanks,

 

[KootK]

I come at it like this:

1) 15% pre-pencil sharpening.

2) 5% post-pencil sharpening.

I like #1 as my reality check as I can usually find a way to tease plausibility out of just about anything if I want it badly enough. #2's for the calc file.

 

[JoshPlumSE]

Really interesting thread. I think every engineering has to come up with their own practice based on their comfort level. Some of my thoughts on this subject:

1) For new construction I try not to go over 1.0. But, if I'm over by less than 1% then I don't stress it at all.

2) For existing construction, where I'm adding new loads to the beam, then I design it to a max of something like 5% over stress.

3) For existing construction, where I'm NOT adding new loads. An example is a situation I've dealt with where I've been tasked with designing / repairing a failed roof truss. The new truss that's replacing the broken one I treat it like new construction where I only let myself go over by less than 1%. However, for the other trusses which are NOT damaged, I will let them be if the code check is less than 50% over stressed. However, I will let the owner / client know about this issue in writing. I may provide a "voluntary" retrofit that would intended to increase the capacity of the truss and/ or reduce the likelihood of damage. I don't worry if this voluntary retrofit has a DCR of 1.5 if it is significantly lower than it would have been before the retrofit.

4) I've revised the load factor for dead load before if I had a good idea what the load was really going to be.... (i.e. hydro-test loading for heavy industrial projects).

5) Alternatively, I do what KootK calls sharpening the pencil. I take a closer look at all values used in the program (RISA in the OPs case) Cb, unbraced length, et cetera. I might take a look at what would happen for moment "re-distribution" were a ductile failure to begin. If I expect that the moment would easily re-distribute to other areas, and the structure would not be compromised by this issue, then I might allow a larger failure. Maybe 25%. As long as there is still capacity left in the beam after load re-distribution has occurred.

 

[Eng.Abdulla95]

Oh, I tend to stick to 0.8-0.9... just because I know those people will do crazy stuff.

Anyway, aside form being too conservative I should point out some of the things that you need to consider (correct me if I am wrong):
* code allowable loads accounts for small dynamical effects. If the activity under consideration is above normal the engineer need to do proper dynamical analysis.
are maintenance equipment have normal dynamical effects or it has large vibrations?
* Site inconvenient and additional eccentricities added from imperfection, can you grantee that the structure will be built 100% as design drawings?
* Redistribution of loads, are you sure that your member will not carry addition loads distributed from other members? How confident are you about your idealized model?

Last words, it may be safe to recalculate code coefficients and factors, but do you expect your structure to preform well for its intended life time, maybe you are just shortening its life span (maintenance period). Please note that the stated webinar above dose mention reliability factor in the equations for safety factors, which means reducing the reliability of the design!! you should not feel good producing less reliable designs from average!

** For honesty I am new engineer

 

[Agent666]

I've always found by playing around with the exact location you are putting whatever you are lifting down, or laying down some additional steel as dunnage or something to spread the load further than the immediate footprint you can probably get things under unity. Also consideration of plastic design and some redistribution of the loads if you are dealing with continuous members can often net you 10-20% more load carrying capacity.

Alternatively just strengthen the members in question before undertaking your maintenance work if the risk is deemed unacceptable.

For the reasons noted in that video, if its a one off loading I'd tend to be more relaxed about it, but get agreement between a few engineers involved to make sure everyones ok with it and understands the risks.

If in doubt you can always take out some of the uncertainty in some of the factors by testing some of the steel from the beams to determine yield strength (assuming here it's structural steel). You'll find you probably get 10% extra right there which for a one off load you are mobilising.

 

[Agent666]

I'd point out as well, in the case of structural steel, you could only justify the overload if it failed in a plastic manner. If it failed by lateral torsional buckling at a lower design moment lower than the full capacity I'd not consider any increase acceptable as its not a stable failure mechanism. But if it was fully restrained such that the full moment capacity was obtained which is a requirement for plastic design, then it's more acceptable based on judgement.

Another option is back propping to some structure below, to share the load over several levels. Quite common when building multistorey structures for example.

 

[Tomfh]

This suggestion to let the client override the requirement - is that truly a valid approach?

Generally they have no engineering training, no understanding of risk.

It all fails the “newspaper test” to me. How would it read the next morning should something go bad.

I look at all these bank foreclosure cases, when people knowingly borrowed more than they could afford, to the extent of LYING to the bank about income, and still it’s the bank which is blamed, because they are seen to have a duty of care to protect people from themselves.

I don’t know any analogous precedents in engineering but I struggle to see how it would stack up in court.



Here in Sydney we’re having some issues with failing apartments. Serious failures of concrete beams, footings, etc.

If anyone got ahold of a design document which showed one of the failed elements was knowingly overstressed, that designer would be in a whole world of pain. A clever journalist would love nothing more than to find a line of computer output that said “NOT OK” but which was passed anyway.

 

[12345abc6ttyui67]

One thing to note (if it's a steel structure). Try getting the mill cert's for the steel that was supplied. You will probably find it has a yield stress a decent chunk over the minimum. I've used this to get me out of dodge a few times where supposed 355MPa steel has come in at 380 to 390 MPa.

 

[ThomasH]

Tomfh
I would say it depends a lot on what type of client it is. Most of the work I am involved with is done for preofessionals. They usually undersstand the concept of risk and how codes will create a certaing margin of safety.
I have heard something similar to you "Newspaper test" before. I don't think it is a bad idea. But I have seen newspaper articles that are so far from the facts that I would not entirely trust the newspaper anyway.

When I wrote my first post the intended approach was similair to a project I was involved in a few years ago.
A structure on a industrial plant had been damaged over time (corrosion etc) and the first question was, is it safe? The second was, can we increase the loading slightly?
The Maintenance Engineer (same role as OP) had concluded that he was not comfortable with his own analysis. He worked with 2D frames and beam/column elements. Nothing wrong with that but under the circumstanses he did not feel that it was reliable. So I was asked to do a "better" analysis (since I have done this more than once before). That meant non-linear 3D FEM-analysis using plate elements to model the critical parts (detailed stress plots etc). We also had a laser scan of the structure to ensure that the geometry was correct. This thing was several meters up, connected to a lager structure.
I agree with RandomTaskkk that the material can be significantly better than stated but in the my case we stared with the assumed material since the structure was old. We could have tested it but "saved" that option.

Somebody might think that this approach is very expensive, and it definetly costs more that designing a new structure. But the cost for this analysis, compared to stopping the plant, demolishing the existing structure and building a new structure, that choice was easy. The end result was that a few members were braced to decrease their buckling length, and the web was reinforced in a few beams because the existing web had corrosion damage. And we were never above 100% utilization and we got a happy client because the production never had to stop.

I am by no means saying that this is always a valid approach. That is why I asked what the consequense of a "no" is.

But I was a bit surprised the AskTooMuch is a Maintenance Engineer, and afraid to get sued. If I understand it correct, that means that his employer can sue him. Well, I am not working in the US (I am in Europe) so there may be a difference. But it was a surprice. If my work towards saving money for a client and ultimatly making my employer look good would result in me being sued, I would definetly think twice .

Thomas

 

[SlideRuleEra]

This suggestion to let the client override the requirement - is that truly a valid approach?

Oh, yes, very much so (but on rare occasions), in fact it was part of my job as an electric utility employee (Generation). Same applied for my mechanical and electrical counterparts.

For the type situation in this thread, whether our Plant Engineer or a Consultant, we would work together to decide what technically detailed design criteria was appropriate. On the infrequent occasions when a consensus could not be reached, the decision was ultimately my call.
Note, "lowest cost" was way down the priority list. Reliability, safety, and cost control were all higher priority.

The OP mentioned plans to ask about what live load would be expected... good move... and I'm the guy he would ask... which would lead to a much more in-depth look at the entire situation.

http://www.SlideRuleEra.net

 

[BridgeSmith]

Thanks ThomasH and SRE for weighing in on my supposition. I definitely agree that coming back to someone with a 'this may or may not fall down if you load it like that' is not the preferred option, nor likely to make anyone particularly happy. However, when it's that or 'nope, you can't do that', clearly communicating the risks and uncertainties to those who will take on the responsibilities for the decision, both safety-wise and financial, seems like it could be the best option.

 

[Tomfh]

I have heard something similar to you "Newspaper test" before. I don't think it is a bad idea. But I have seen newspaper articles that are so far from the facts that I would not entirely trust the newspaper anyway.

It's a good test, although a rather stringent one!

Let's say the beam collapses (for whatever reason), and someone is hurt, and you designed the beam, and your computer model showed it as red "NOT OK". That will look bad in a newspaper. The journalist (or judge) can point to the red "NOT OK" beam, and you will look very bad.

When it all goes well and you save money, that's great. Everyone's happy. Client saves money. You're the clever engineer who saves them money. I'm just talking about the rare times when it doesn't work out.

 

[ThomasH]

Tomfh
Like I said before, I like the test as a mental exercise. But I would not call it "stringent", I prefer the word "random" because my experience is that journalists have a rather random ability to get the facts straight. I don't think they lie on purpose but they want to sell newspapers and facts have a tendency to not be very dramatic. And in some cases the facts may be confidential.

I have been in that situation myself. But it that case the issue was not a failure but discussions regarding comfort in a high building. The fact was that there was no problem.

Thomas