Structural Calculation Standard of Care 1

[Gumpmaster]

Is there a standard, somewhere, which says what should be included in structural calculations?

For instance, if I have a a deferred submittal where the calculations are done by the Contractor's engineer, what should I expect? I received a set of calculations that calculates all the loads for a paritcular component, but doesn't calculate any capacities and doesn't compare the applied loads to anything that would show that the component is acceptable. I know as the EOR we don't need to do a line by line calc check but we typically check for gross errors.

The problem I have is that the Contractor says that there's nothing requiring them to compare applied loads to capacities. Is there any standard which says what should generally be included in calculations?

 

[Archie264]

If you don't have the ability to reject something then there's no point in them submitting them. That should be rejected right away to let the contractor know it's not acceptable. This is something worth asserting yourself on, in my opinion, and the sooner the better...professionally, of course, but that needs to be nipped in the bud.

As for your question itself, well, that's the whole point of a safety factor, or resistance factors, if you will. He's given you the load factor, presumably, so ask him for the resistance factors. Otherwise it doesn't comply with AISC, ACI, NDS, etc.

 

[bpstruct]

Sounds like the contractor's engineer has provided you with nothing. Not sure what the component is, but most codes and material standards require phiRn > Ru (in some form or fashion). Or the calculated stress must be lower than the allowable stress. Even without a specific code to refer to, common sense should trump here anyway.

I would reject the submittal.

 

[Gumpmaster]

Yes, I rejected it. 2nd go round the Contractor says it's a change order and is saying their work may slow down as a result. I'm rejecting it again, but fights like this don't benefit anyone.

This isn't the first time we've run into this though. Quite a few times we've had people say that the analysis is required but they're not required to document it in the calculations. We typically require "full and complete calculations." We reject and they refuse and it often takes weeks (if not months) of back and forth before they finally comply.

It would be nice to be able to point to something concrete in response to the Contractor. The closest thing I've come up with is IBC Section 1604, General requirements. It says that the design should be in accordance with LRFD or ASD, but doesn't require documentation of that.

 

[Ron]

In many US states, this would be considered a "delegated engineering function"; whereby the structural engineer of record sets the criteria and the delegate engineer (in your case, the subcontractor's engineer)has to meet the criteria. Nothing in this process should require a change order. While in most instances the SEOR is required to provide the criteria in writing, there are instances where the plans and specs provide such criteria without a separate document. In that case the delegate engineer should provide you, as the SEOR, with adequate proof of capacity through independent calculations, tabular product data or other info. If that is not done, you should reject it as you have done.

The contractor is using a typical ploy of extortion to get you to drop the requirements. Don't! (and it is apparent that you won't)
He's making you be the bad guy....no problem. It's apparently our job.

I've done hundreds of delegated engineering jobs and I've had the SEOR come back to me a couple of times asking for more info. I have no problem with that and provide it if possible. (I've also had a couple of engineers ask for some ludicrous things as well....not often, but has happened. For those I usually object if I've provided sufficient information for them to determine adequacy in accordance with the stated project requirements).

 

[oldrunner]

To prevent instances like this, there should be specific clauses in the contract specifications. Then if the contractor doesn't comply, he is in violation of the contract. Sometimes we find that we can turn a submital from the contractor around in about two weeks. Then there's the hassle like you are going through presently, which might take 6 months. Or there is the instance where it's taken a year or maybe never for approval. There is some sections in the Blue Book which somewhat address your problem.

 

[Gumpmaster]

Thanks, I may have to check out what the Blue Book says.

I've considered changing our specifications, but it's a shame to change specifications each time you have a yahoo that screws something up. I may end up doing it though. We have specs like that because one time 20 years ago such and such happened and we don't want it to happen again. You end up with a complicated set of specs. There was a time when everyone on a project tried to do the job to the best of their abilities with the final outcome in mind. It seems those times may be past. There was also a time when you could build a project with a minimal set of plans and nearly no specs on a handshake. That time is definately past.

 

[JStephen]

In the past, I've used some structural software where the basic software did the analysis, which meant it calculated all the moments and forces, but it didn't do a code check. That was a separate package, or if you needed to show compliance with a non-standard code, it involved a lot of handwork or spreadsheet analysis. That might be what you're running into. In any case, the analysis itself is pointless if you can't show that the calculated results are acceptable. (It is possible in various circumstances to calculate low stresses that are unacceptable due to buckling or stresses in excess of yield that are still acceptable.)

You may have instances where the delegated engineer designs an item, and then gives you the loads that it applies to the structure you're designing. In a case like that, they wouldn't normally be checking your structure for adequacy, just theirs. I'm assuming this is not your situation.

Typically, when designing as a delegated engineer, we'll generate calculations if requested to do so, and those will include major items. But not necessarily every single calculation that could be calculated, either. Part of the responsibility of being an engineer is knowing what needs to be calculated and what doesn't. But in these cases, we can and do generate additional calculations on specific items if requested.

 

[msquared48]

No written standard I know of other than the contractural requirements with your client, public safety, the law, and your professionalism. Go get 'em... Ron said it well.

Mike McCann, PE, SE (WA)

 

[steellion]

Your specifications should dictate that calculations for delegated designs should be signed and sealed by a P.E. in the state having jurisdiction. Contractors' engineers are much less likely to give sloppy, incomplete, or erroneous calculations when their stamp is on the page.

 

[JedClampett]

I've started avoiding asking for calculations for all the reasons stated above. We used to ask for open web joist calculations. I started just picking the joists out of the SJI Standard because having the manufacturer design them and provide calculations was worthless. For wood trusses, I still ask for them, but they're also indecipherable.
Getting a pdf of a bunch of computer vomit isn't useful. I strongly suspect that this is intentional. "Screw them, if they want calculations, I'll give them calculations that would choke an elephant!" seems to be the attitude. No summary, no assumptions, no applicable loads, basically, some DOS based programs from the early 80's in 6 font that vaguely look like the element in question.
Sorry, venting.

 

[TLHS]

I had a great one where a supplier was required to have an item engineered and provide calculations upon request. The drawings we got were terrible and a spot check showed failure under expected loads, so we asked for the calcs. They sent us calcs that didn't match the arrangement, showed poor understanding of the forces and had no load path. We pointed out the issues and asked, at minimum, to get calcs that actually matched the arrangement on the drawings.

The contractor had us talk directly to their engineer, who maintained that legally he didn't have to provide anyone with calcs and his stamp was assurance. I agreed that he legally didn't owe *me* anything, but told him to talk to his client since they had a contractual obligation to provide calculations. So then I got to have fun conversations with the supplier about how their internal contractual situation with their contracted engineer was their problem, but they had agreed to supply certain things to my client.

There was lots of fighting, and I never did get acceptable calcs. We basically had to re-engineer what was supposed to be delegated work to ensure that it would be adequate.

It never made sense, since it was maybe two hours worth of math. The supplier and their engineer spent way more than two hours arguing about it and sending math that didn't appear to have any relation to what was being checked.

But yeah, I hate delegating engineering on anything we could reasonably do internally.

 

[Archie264]

Getting a pdf of a bunch of computer vomit isn't useful. I strongly suspect that this is intentional. "Screw them, if they want calculations, I'll give them calculations that would choke an elephant!" seems to be the attitude. No summary, no assumptions, no applicable loads, basically, some DOS based programs from the early 80's in 6 font that vaguely look like the element in question.

Yep, definitely been-there-received-that.

When I get said pdf emailed to me the first thing I do is hit <ctrl> f (i.e. "find", i.e. the search function) and type in the work "fail". Much of the time the calcs they submitted show that their own design failed. It would be concerning if their model actually resembled reality in the first place. :-D

The dos based software brings to mind software for mat footing design. It's amazing how long that one's been used.

 

[charliealphabravo]

So what triggered the requirement for calculations? As the EOR I'm not usually asked for them and I don't usually ask the delegated engineer to supply them. It seems to me that receiving and approving calculations attracts some additional degree of liability for the EOR. Some building jurisdictions I have submitted work to felt the same way and they would rarely request calculations for this reason.

 

[Gumpmaster]

Well, as the EOR we want to make sure that the building is safe. We've had experiences similar to TLHS and Archie where submitted items can't calc out. Had they been installed, they may have been unsafe for the public. I don't think there's anything that legally requires us to ask for calcs, but it's a comfort thing for me. Also, depending on the jurisdiction, you may need calcs to satisfy the permit. The difficulty is that no one dictates the content of those calcs so you may get something good, you may get something half hearted, and you may get computer vomit.

 

[racookpe1978]

I'm getting more an more suspicious that they (the original contract team and their arrangement dwgs never did the calc's, nor specifically wanted to "declare" the assumed values for the actual loads and forces - if known at all to any degree of accuracy), and want you to take the responsibility for subsequent failures.

 

[charliealphabravo]

I think we are all on the same page here. I am not completely against presenting/reviewing calculations although I do passively/aggressively resist on occasion. I think my point is that I don't want to be responsible for something that I haven't been paid to be responsible for. The delegated engineering process is specifically designed to draw clear lines of responsibility.

If the delegated engineer has been negligent then you cannot entirely remedy that problem by reviewing calculations. You can surely mitigate the problem and possibly facilitate the negligence of the delegated engineer on the current and future projects.

 

[graybeach]

I work on the owner's side, and my RFPs have evolved over the years to require EORs to submit their calculations and to review delegating design calculations. In both cases, I require that calculations show evidence of being checked by someone other than the person who prepared them. When I only said that calculations shall be checked, often the same person would sign them as both the originator and the checker.

Even with all of this, EORs balk and often miss things so I still have to check both the EOR's and the delegated calculations, and it is not unusual for me to find serious errors. If something fails, it is not as if I can just pass all of the consequences to the EOR and delegated designer and go on my merry way. My company is a power utility and will definitely pay a price whether in lost generation, legal fees, or liability.

When I worked for a consultant I was very conscientious, as it seems are most people on this forum. I had no idea how many consultants give short shrift to their work.

 

[JedClampett]

Graybeach, even though I'm pretty hard on the calculation factory approach, I'm not sure anyone goes into the business with the idea of doing sloppy or incomplete calculations. I've never been in the small consultant business, but I suspect that these guys are paid peanuts when they're paid at all. When you're getting $200 (or whatever) for a package of truss calculations, you have to do a lot of them to pay expenses. There's no way that quality doesn't suffer.
My gripe is the presentation. I don't think it costs more to state assumptions, give loads, define materials, etc. than just blowing and going. They're probably the same on every project. And hasn't anyone updated PEMB and wood truss software in 25 years? It would have to be cheaper in the long run to buy better software that gives a decent 21st century product than live with crap that isn't supported anymore.