Contractor Designed Elements 7

[EDub24]

Hi, I've got a general question. Typically I push some items that require design onto the Contractor (guardrails, metal stairs, pipe supports etc). They usually have an in-house Engineer design/stamp the items or they sub out to another engineering firm. My question is how in-depth of a review do you typically do when the Contractor submits the design? Where I work I've seen 3 trains of thought.

1) It's the Contractor's design/stamp so they are only submitting for the record so no review is done.
2) A thorough in-depth review where everything is looked at and verified/checked.
3) A cursory review where the design criteria, design procedure and some select structural items are looked at.

Typically I do #3 and perform a cursory review unless the designed element is pretty major in which case I'll do a more thorough review although in both cases I'll word my comments as suggestions/requests to verify or check rather than direct them to do something. Thanks!

 

[JedClampett]

I tend toward number 3, with a check of a few places where I suspect I would of had problems. For instance, if we delegated metal stairs, the connection at the top is always an issue. My opinion of methods 1 and 2 are:
[ul]
[li]If you think you're avoiding liability by just filing them without looking at them, you're not. I can just see the lawyer asking me, "So you were sent these calculations and you didn't even look to see that they were in meters, not feet?"[/li]
[li]If you do a thorough number by number review, you might of well of done the design in the first place. Plus, you're just inviting a pissing contest between you and the delegated engineer. He sealed it, he's assumed competent, why do you need that intrusive of a review?[/li]
[/ul]

 

[EDub24]

I agree with both of your opinions on methods 1 and 2. For the 1st method, one of my coworkers told me that someone said that to them and my jaw dropped.

 

[BridgeSmith]

When we review designs that someone else stamped, we generally look for things outside of the design itself. Things like:

Has the correct loading has been applied to meet the project requirements?
Have they checked constructibility and stresses on components during construction?
Are there interference issues with other components of the project?

 

[Ron]

EDub24....this is known as a delegated engineering function and somewhat varies with State law in the US. The basic requirement is the following:

1. As the design structural engineer of record, you establish the criteria to be used by the delegated engineer, even if it is only to follow the prescriptive criteria of the building code. You establish the loading if components and cladding or require the prescriptive loading as necessary (wind, seismic, etc.)

2. You must then review with enough detail to determine if your criteria were followed and the results are reasonable. You don't have to do an exhaustive review or re-create anything, but do a reasonable check for compliance to your established criteria. You should document your review and your "acceptance" of the engineering submittal (the common term for pseudo-acceptance is "No Exception Taken")

This is the standard of care.

 

[TehMightyEngineer]

I'm the in-house structural engineer for our precast company but have worked in consulting firms for most of my career so I've seen both sides.

I highly lean towards #3 being the best option. Speaking from the point of view of a delegated design engineer, we've seen more projects get derailed when option #2 was used and fewer projects derail when option #1 or #3 were used. But there is a trade-off, use delegated design in a common-sense manner and you'll have the best result.

When you delegate engineering have a robust specification that clearly describes the delegated design and the performance criteria required to be designed to. I've seen far too many specifications that were boiler plate and clearly just cut and pasted from other jobs. We just finished the design for a simple precast pedestrian bridge that ran over schedule because the specification was written for a steel bridge. You only open yourself up for liability and raise the cost of the project by having a poor spec.

Simultaneously, don't make the specification overly restrictive. Very commonly I see specifications that restrict me to certain design methods, materials, loadings, and so on. This can be good if applied with thought to ensure a properly designed product but can easily do nothing but raise the cost with no benefit. I've designed concrete covers for heavy vehicle loads when the cover was raised 1 foot above grade with bollards and equipment on top and no feasible way to get a truck there except if they crashed down a hill. I've had to design with extremely high grade stainless steel rebar because the design engineer specified exactly that grade had to be used; despite other equivalent stainless steel grades being more readily available for much cheaper.

In short, specify exactly what you need for the project to succeed and no more. Let the delegated engineer actually engineer; if you force me into one path I'm going to take that path and be forced to skip anything that saves time or money.

All that said, review sufficiently to ensure the delegated engineer is doing a proper job. It does me no favors if you allow a poor design through. It sucks if I'm competing against someone who doesn't even have an in-house engineer and just keeps going off of the same design for a product they had done 5 years ago while I'm preparing detailed designs for each job. Make sure the playing field is level and that you're getting the quality of product you or the owner wants/needs. Remember that if the product fails the owner will likely drag the EOR into whatever fallout occurs. Leave yourself as much room as possible to put the blame on the delegated engineer because I'll happily throw that blame on the EOR if I reasonably can.

Often I'll see reviewers who are charging for the review and/or have to justify their review process. I actually often leave one or two minor design "errors" (that don't affect the design of course) in many submittals for them to "catch". We've had one reviewer go so far as to call out spelling errors because they couldn't find anything else to correct, but didn't want to approve the submittal without at least one round of revision and resubmittals. Please don't do this, I've seen way too many dirty things done by reviewers who were charging for their review. I'd love to detail some of the more egregious experiences with dirty reviewers trying to rack up billable hours but I must protect the innocent and avoid pointing fingers.

Regarding submittals for review, only ask for things you actually need to review or document. If you ask me for submittals on every last design detail I will happily give you 200 pages of RISA-3D print outs and spreadsheet calculations and double the price of everything. Preparing these calculation submittals takes a lot of time and the more we submit the more we assume you'll find some irrelevant detail to return the submittals on delaying the project. I've found that if I submit two identical projects; if I submit one with only the bare minimum of calculations and supporting documents, and one with full calcs, figures, load diagrams, etc.; the more complete calculations will take longer for me to get returned and often have more comments on items to "revise and resubmit". In short, be aware that we will charge more money for the product the more onerous the submittal requirements are.

By far the best projects are always the ones where the delegated engineer and the project engineer of record work collaboratively to ensure project success. I'm working right now on a DOT project where we found a design omission on the EORs foundation for our precast component and we worked with the DOT and EOR to provide a precast concrete foundation solution that not only addresses the issue but improves the expected project timeline and overall costs. Everyone wins. The EOR should be watching the back of the delegated engineer to catch any mistakes that could derail the project. Simultaneously the delegated engineer can and should bring up anything the note during their design process that might help out the EOR and avoid similar derails of the project.

Jed makes great points from the EOR's point of view.

If you do delegate engineering understand what you're delegating and why. Don't delegate to avoid the work and essentially make the owner pay twice for the engineering. I've seen far too many engineers delegate items that should have been part of their design but because it was difficult or time consuming they simply put a box on their drawing and label it "by others". Do delegate engineering if it is a specialized product (like precast) or is a field not in your area of expertise.

Ian Riley, PE, SE
Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[dik]

From my Project Notes:

SHOP DRAWINGS AND SAMPLES

CONTRACTOR SHALL PREPARE A LIST OF SUBMISSIONS

REFER TO GENERAL NOTES FOR PARTS OF THE WORK THAT REQUIRE SHOP DRAWINGS/SAMPLES. ALLOW 2 WEEKS FOR REVIEW FOR EACH COMPONENT U/N

SUBMIT SHOP DRAWINGS/SAMPLES WELL IN ADVANCE OF DELIVERY

SUBMIT SHOP DRAWINGS/SAMPLES TO THE [OWNER | CONSULTANT | ENGINEER]

SHOP DRAWINGS SHALL BE SUBMITTED AS THEY ARE PREPARED. UNLESS PRIOR ARRANGEMENTS ARE MADE, DUE TO SCHEDULING DEMANDS, ETC. THE [OWNER | CONSULTANT | ENGINEER] CANNOT BE HELD RESPONSIBLE FOR DELAYS CAUSED BY THE RECEIPT A LARGE NUMBER OF SHOP DRAWINGS IN A SHORT TIME PERIOD

CONTRACTOR SHALL ALLOW AND SCHEDULE FOR TWO WEEKS FOR [OWNER | CONSULTANT | ENGINEER] TO REVIEW EACH SUBMITTED SHOP DRAWING/SAMPLE

ELEVATIONS AND DIMENSIONS USED SHALL BE THE SAME AS THE CONSTRUCTION DOCUMENTS U/N

SHOP DRAWINGS SHALL SHOW ALL MATERIALS AND DESIGN, FABRICATION, CONSTRUCTION, FASTENING, AND FINISHING DETAILS AS REQD

SHOP DRAWINGS SHALL IDENTIFY ALL VARIATIONS FROM THE CONSTRUCTION DOCUMENTS

THE CONTRACTOR SHALL SUBMIT 1 SET OF SHOP DRAWINGS IN DIGITAL FORMAT PRIOR TO FABRICATION FOR THE [OWNER | CONSULTANT | ENGINEER]'S REVIEW. SUBMIT SHOP DRAWINGS FOR THE FOLLOWING WORK U/N:
-CHAIN LINK FENCING
-COLD FORMED METAL FRAMING
-CONCRETE FORMWORK
-CONCRETE POUR SCHEDULE INCLUDING ALL CONSTRUCTION/CONTROL JOINTS
-FACTORY FABRICATED WOOD TRUSSES
-FIBRE REINFORCED PLASTIC COMPONENTS
-LONG SPAN STEEL TRUSS
-METAL FABRICATIONS
-METAL SIDING
-OPEN WEB STEEL JOISTS
-PRECAST CONCRETE ELEMENTS
-PRECAST CONCRETE PAVERS
-REINFORCING STEEL
-ROOFING
-STEEL BUILDING SYSTEM (ALL COMPONENTS)
-STEEL FLOOR DECK
-STEEL ROOF DECK
-STRUCTURAL STEEL
-WOOD GLUED LAMINATED PRODUCTS
-WOOD 'I' JOISTS

PRIOR TO REVIEW BY OUR OFFICE, SHOP DRAWINGS/SAMPLES MUST BE REVIEWED AND COORDINATED BY THE CONTRACTOR. THEY MUST BEAR A REVIEW STAMP, DATE AND SIGNATURE SIGNIFYING HIS ACKNOWLEDGEMENT THAT HIS RESPONSIBILITIES INCLUDE:
-DIMENSIONS WHICH SHALL BE CONFIRMED AND COORDINATED WITH JOB SITE CONDITIONS,
-INFORMATION THAT PERTAINS SOLELY TO FABRICATION PROCESSES, MEANS, METHODS, TECHNIQUES, SEQUENCES AND PROCEDURES OF CONSTRUCTION,
-SAFETY PRECAUTIONS AND PROGRAMS IN CONNECTION WITH THE WORK,
-COORDINATION OF THE WORK WITH THAT OF ALL OTHER TRADES
-SATISFACTORY PERFORMANCE OF THE WORK

SHOP DRAWINGS NOT BEARING A REVIEW STAMP IN ACCORDANCE WITH THE ABOVE MAY BE RETURNED

SHOP DRAWINGS FOR NON-STRUCTURAL COMPONENTS WHICH MAY IMPACT THE BUILDING STRUCTURE SHALL BE REVIEWED BY THE [OWNER | CONSULTANT | ENGINEER]

SHOP DRAWING/SAMPLE REVIEW BY THE [OWNER | CONSULTANT | ENGINEER] IS ONLY FOR THE LIMITED PURPOSE OF CHECKING FOR CONFORMANCE WITH THE INFORMATION GIVEN IN THE CONSTRUCTION DOCUMENTS. REVIEW IS NOT CONDUCTED FOR THE PURPOSE OF DETERMINING THE ACCURACY OR COMPLETENESS SUCH AS DIMENSIONS OR QUANTITIES

These are modified to suit the project with some additions and some deletions. I stipulate in the various sections what shop drawings require engineering seal.

Dik

 

[fel3]

I tend toward #3 also, and only perform #2 if I spot significant problems. I have noticed that every time I needed to revert to #2, over my 38-year career, it was because (as I found out later), the delegated engineer was a young PE with little or no effective supervision or mentoring AND they had performed the calculations by hand or in Mathcad, so it was easy to spot the problems. With spreadsheets and computer printouts, it can be more difficult.

Most of the calcs I have reviewed over the years have been for buried traffic-rated pre-cast vaults, but I have also reviewed calcs for AWWA D-100 welded steel water tanks, overturning and anchorage calcs for polyethylene and fiberglass tanks, equipment pads, hydraulic and other calcs related to pumps, and others.

Examples of problems that have caused me to perform #2:
– Results that are vastly different from what I usually see.
- Calculations without units or with few units. I have seen this with hand calcs and, strangely enough, with Mathcad calcs. In each such case, I found significant problems with some of the results because the needed unit conversions were handled incorrectly, which is easy to do if you don't write them down.
- Dimensions used in the calculations that do not match the diagrams, sometimes not even closely. On the other hand, I can't recall reviewing a calculation with incorrect loading conditions, so at least they got the code part right.
- Mixing lbm and lbf in Mathcad, resulting in numbers with incorrect units.
- Screwing up the units in empirical equations. And, in Mathcad, performing manual unit conversions instead of letting Mathcad handle it.
- Taking the result from a calculation at the bottom of a page in Mathcad and redefining the variable with the previous result (but with a typo) on the top of the next page.

==========
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

 

[TehMightyEngineer]

fel3: Forgive me if I derail the topic but maybe you can tell me.

Why do I get so many requests for signed and sealed buoyancy calculations for tanks? I've seen enough jobs that required this especially to notice a pattern. Is it simply because so many contractors/non-engineers get this calculation wrong?

It's one of those items that I understand why it's delegated but don't understand why the submittals are so onerous.

Ian Riley, PE, SE
Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[dik]

Mixing lbm and lbf in Mathcad

That one of the reasons I like SMath... at the beginning I define: 'lb' as 'lbf', 'K' as 'kip', 'ft_K' as 'ft*kip', etc...

Dik

 

[TehMightyEngineer]

Oh, a BIG note I forgot; do not specify anything that could be insufficient to meet code or unconservative!

I've had a delegated design where the EOR specified loading criteria that resulted in an unconservative load case (soil load counteracting a gravity load). We did the correct thing and enveloped the design with their load case and our conservative load case but we could have easily designed for only as written in the specification and had an easy way to point liability onto the EOR if it somehow failed.

If you tell me to design precast floor panels to 60 PSF live load, but ASCE 7 and the IBC require 100 PSF; if I design to 60 PSF as required and it fails I definitely believe it would be the EOR who is liable for the design error if that's what was specified.

As a practical nature, if I think a specification is unconservative or not sufficient for a design I will either discuss it with the EOR or "overdesign" a product. Other delegated engineers might not be so helpful or thorough.

Ian Riley, PE, SE
Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[MrHershey]

#3 with some of #2 if it seems like they don't know what they're doing. But really try to stay in #3 as much as I can. I'll also usually look up their license and make sure it's current and in good standing.

Also try to treat it more as a peer review. Check basic design criteria and then a brief review with some spot checks. Try to keep to comments like 'hey did you look at this' or 'I noticed this, please verify' instead of 'this is wrong, please fix'. The other person is still an engineer, usually has the exact same license I do. So they deserve the same courtesy I would hope to be provided were I in their shoes. Everybody has different ways of doing things, I don't want to be dictating to some other engineer how to do his or her work (unless they're wrong or we have real specific requirements).

Often I'll see reviewers who are charging for the review and/or have to justify their review process. I actually often leave one or two minor design "errors" (that don't affect the design of course) in many submittals for them to "catch".

Someone else using my tricks. Don't think I've done this as a delegated engineer, but have done it as EOR. Will do this primarily on jobs where I know reviewers will keep digging until they find something to comment on (military is usually the biggest culprit). Usually won't leave design errors, per se. Instead will leave out minor information that would then elicit an RFI somewhere down the line like leaving obvious placeholders for loads/requirements for delegated design items or not calling out the size for something.

If you do delegate engineering understand what you're delegating and why. Don't delegate to avoid the work and essentially make the owner pay twice for the engineering. I've seen far too many engineers delegate items that should have been part of their design but because it was difficult or time consuming they simply put a box on their drawing and label it "by others". Do delegate engineering if it is a specialized product (like precast) or is a field not in your area of expertise.

Agreed. Try to not delegate critical items. Only times we really delegate are when it's a speciality product (precast, FRP, joists and joist girders, some foundation systems, etc.) or something that someone else can do a lot cheaper than we can, so us omitting it from our scope/fee and delegating actually saves the owner money (pre-engineered buildings, pre-engineered stairs, non-load-bearing light gage framing, steel connections in some parts of the country, etc.)

 

[JedClampett]

As long as we're getting off topic, I'll embellish what MrHershey concludes. You should only delegate items where the contractor or supplier has better knowledge of the design or the materials.
Anyone could run the steel tank wall thickness calculations. But there's a good chance of a cost savings if they can pick the steel grades and thicknesses (I found out that they can get steel rolled to 100ths of an inch just to meet design), so they expect to do their own design. Same with prestressed tanks. FRP members are another one, where there's guys (and gals) who know a lot more about it than me.
On the other hand, we delegate grating designs in our master specifications. Well, if I can't do my own grating design, I should be retired. It's s simple table to get the grating depth, plus I like a beefy grating. So I take that out.
I think over-delegating design is s chickenpoop way of doing work. I just delays the pain. Even if you have an inadequate design budget, sooner or later, you're going to have to pay the piper.

 

[TehMightyEngineer]

Usually won't leave design errors, per se. Instead will leave out minor information that would then elicit an RFI somewhere down the line like leaving obvious placeholders for loads/requirements for delegated design items or not calling out the size for something.

Yeah, I should specify that I don't really introduce true errors. Usually I'll do something like incorrectly label something (like call the north wall a south wall), basically anything that the reviewer should easily find but is easily addressed in an RFI and has no impact on design. Anything I feel people would need to rely on to follow my calculations will always be as correct as possible.

Ian Riley, PE, SE
Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[MrHershey]

so they expect to do their own design

Have increasingly been seeing this with the cold-formed design. We try and get out of it whenever we can but some owners or archtiects fight to keep it in our scope. Have had a couple jobs where that happened, we designed the exterior studs for wind loading, and then the contractor came in with their sub who threw our stuff in the trash and engineered everything themselves (which I was certainly happy to accept as an alternate). At the time (maybe still) some of the cold-formed companies were throwing in engineering and drawings for free if you used their studs, so could have been absolutely zero cost to the owner but they insisted on paying for it.

 

[NorthCivil]

In the frozen north, I've worked in both desks, and never checked anyone's secondary structure calculations or had anyone check mine.

Even working on the big dollar projects with the big global players, I've never had anyone ask me for my calculations. Though I have carried out general concept reviews of the drawings only and had that carried out on my drawings as well. Different ball game up here maybe?

More often than not, there has been little to no interaction between myself and the other engineer at all.

 

[fel3]

TehMightyEngineer: I don't know why you are having to do the buoyuancy calcs, though I suspect you are correct that the design engineer may not know how to do them. But, it's not like the method is a mystery. Or is it? Or, perhaps the engineer just wants your confirmation of his calcs.

I have done my own buoyancy calcs for numerous buried vaults and lift stations and I have never (to my recollection) asked the delegated engineer to do them, although a few have on their own initiative. I base my initial calcs on my best estimate of the delegated engineer's final structure design, then I revisit my calcs if the structure ends up being different enough (i.e. lighter enough) to matter. IIRC, this has only happened to me one time.

dik: I use SMath from time to time, but overall I prefer Mathcad. Mathcad (and I think SMath as well) already has the built-in units "lb" and "lbm" for pounds-mass and "lbf" for pounds-force. So, when people use them incorrectly, I immediately know that they don't know the program very well. BTW, when I do hand calcs or annotate an Excel calc, I use "lbm" and "lbf". In the late 1970s and well before Mathcad, one of my college professors insisted on it and it became a habit.

Fred

==========
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

 

[JStephen]

From the contractor side, we see it all ways on the design and the submittals.

The major issue I see with delegated engineering is that we'll have an item fully detailed on the contract drawings, with a PE seal, but then elsewhere, there's a clause requiring us to be responsible for the design. About half the time, it's intended one way, half the time intended the other way, and you usually get the impression that nobody else asked the question, either. So I assume there's been a few times where two different engineers each assumed the other was responsible for that item, and something put on the drawings as a "typical detail" was assumed to be the intended design when it wasn't.

 

[dik]

fel3: lb is a unit for lbm... Just checked... I knew there was a reason to 'overwrite' it.

Dik

 

[JedClampett]

Have increasingly been seeing this with the cold-formed design.

I don't understand. Are they redesigning it to use lighter sections? Or are they replicating your design? If they're not saving him money, I don't see what the advantage is for the contractor. I've got light gage metal submittals that seemed like they used gages that you could see through. And those were cases where I provided a design.
This sounds like a race to the bottom. Whichever supplier can provide the lightest design wins the project as long as they can provide some calculations covering his numbers.