How to describe a good design

[enginerding]

We have had a problem with a manufacturer providing very poor quality drawings and documentation for a component to be added to a structure we have designed. It seems that simply stating that the design must meet the 2006 IBC is not good enough to get us a good design.

I am trying to come up with a way of describing a good design that is well documented and detailed so that we can put into writing what we are looking for as far as a submittal. The problem I am having is that it just seems to be more of a gut feel than an easy definition.

We have even been having difficulty getting the design of the component in one package. It seems that we have several smaller components of the system designed by their engineers, but no one is reviewing the system as a whole and putting all the smaller components into a single package. Global stability of the component is an issue that they seem to have difficulty understanding.

Does anyone have a good description of what comprises a good, complete design?

 

[rb1957]

i think you're skrewed ...i think your vendor has been given a contract "design to 2006 IBC" which is what they've done.

i think you wanted to have design reviews (to have input on the design) and/or design acceptance of the component. messily you could have said "design with a view to good practices" but that's open to interpretation.

maybe go over their calcs to verify that the design is compliant, maybe there are a bunch of assumptions that you don't like, then (unless you're mgmt are on side) expect a fight.

good luck !

 

[Teguci]

Tell them to provide a 20 year warrantee

 

[enginerding]

Actually, there was not a clearly defined contract... I am kind of stuck in the middle here, as you seem to have noticed. There were many red flags pointed out, but the system was used regardless. Upon (attempt of) installation, the component collapsed on itself - no doubt due to the very same problems I pointed out.

Money has been paid for the product and whatever agreements do exist require that this manufacturer must be used for this specific project. I have the opportunity now to do better with describing the requirements. Any help would be greatly appreciated.

This is also somewhat of a non-project-specific exercise... I am interested to know if I live too much in my own ideal world and expect structural designs to meet high standards.

 

[Lion06]

What kind of language did you have in the spec. We typically have a performance spec for things like that (metal studs, stairs, things that we don't design). We put language in the spec that we need to see detailed shop drawings, and calculations that show the system can withstand loading per IBC (or a specific load that we want). The calcs and shops need to be stamped by a PE registered in the state of fill in the blank. We also specify deflection limits for things like metal studs.

 

[enginerding]

That is exactly what was stated. What they gave us is a pile of drawings and calculations signed by different engineers for each component going into the final system. There was not a single engineer taking responsibility for the system as a whole that was purchased as a unit. Because of the lack of coordination (or the lack of an overall coordinator) the system was disjointed (overall stability was not considered) and did not act as a unit.

Because this, theoretically, may not be a problem simply for this project, I want to be able to "fool-proof" our spec a little better. A good description of what comprises a good design would be helpful.

 

[rb1957]

"no clearly defined contract" ... either you're completely skrewed if they won't do more work without more money (presumably you don't want to pay them to fix their "obvious" mistakes) or if they're reasonable (and will fix the design) then describe the design the way you think it should be ... you know how it failed, you know what doesn't look right ... how can they modify the design to make it right (even if it looks ugly).

 

[Lion06]

I would send the drawings/calcs back with all of your notes and say that they "have been reviewed for conformance to the design criteria specified in the contract documents, and for compatibility with the design of the primary structure. With regard to these reviewed items, please see the individual sheets for comments. These comments should be addressed and resubmitted for subsequent review." Then, in addition to whatever specific comments you have, I would say on the first page of the submittal, "Please provide calculations for overal stability/strength of entire assembly.".

 

[Lion06]

They clearly haven't met their responsibility of designing for the specified loading or it wouldn't have failed.

 

[H57]

Don't know the exact language to use, but I would try to move down the path of requiring a complete design stamped by ONE engineer. This will require one individual to take responsibility for all the pieces together.

It sounds similar to a situation I have been involved in.
I was asked about some modular buildings that had components designed and engineered by different companies/engineers. Each made assumptions in their design and some of the assumptions conflicted with each other. While each component was engineered adequately (based on their assumptions) the system as a whole did not work. In the end the contractor had to make retroactive repairs and our recommendation to the client was that in the future they should require one structural engineer of record that takes responsibility for the design as a whole.

 

[msquared48]

enginerding:

If you have E & O insurance, you might contact your carrier or agent and get their spin on the question.

Mike McCann
MMC Engineering

 

[BAretired]

Insist on one engineer taking responsibility for the entire package. Tell them their shoddy drawings are not acceptable and are not approved.

BA

 

[Bribyk]

This is where the boss makes (or supposedly makes) his six- or seven-figure income and if he's not doing it well, this is where lawyers make their six- or seven-figure incomes.

You can go at this two ways... give detailed performance specifications (although somethings, like appearance, are tough to quantify) or give detail specifications, which is basically doing the design yourself and asking someone to supply the parts.

"Code" is usually a minimum and generally doesn't cut it for a spec. Basically, if you don't know exactly what you want, you're not going to get it.

 

[enginerding]

I ended up insisting on one engineer acting as EOR of the component to prepare and seal the package, taking responsibility for the selection and approval of the sub-assemblies and for the stability of the system as a whole. I think if that were done, we would be more likely to have a good design.

And code is all I want. I ended up citing a bunch of lines from Section 1603 and 1604 of 2006 IBC. Subsection 1604.4 actually has a pretty good definition of what makes a good analysis:
"Load effects on structural members and their connections shall be determined by methods of structural analysis that take into account equilibrium, general stability, geometric compatibility and both short- and long-term material properties.

"Members that tend to accumulate residual deformations under repeated service loads shall have included in their analysis the added eccentricities expected to occur during their service life.

Any system or method of construction to be used shall be based on a rational analysis in accordance with well-established principles of mechanics. Such analysis shall result in a system that provides a complete load path capable of transferring loads from their point of origin to the load-resisting elements.

"The total lateral force shall be distributed to the various vertical elements of the lateral-force-resisting system in proportion to their rigidities, considering the rigidity of the horizontal bracing system or diaphragm. Rigid elements assumed not to be a part of the lateral-force-resisting system are permitted to be incorporated into buildings provided their effect on the action of the system is considered and provided for in the design. Except where diaphragms are flexible, or are permitted to be analyzed as flexible, provisions shall be made for the increased forces induced on resisting elements of the structural system resulting from torsion due to eccentricity between the center of application of the lateral forces and the center of rigidity of the lateral-force-resisting system.

"Every structure shall be designed to resist the overturning effects caused by the lateral forces specified in this chapter. See Section 1609 for wind loads, Section 1610 for lateral soil loads and Section 1613 for earthquake loads."

Thanks for all your input.

 

[JStephen]

The description is rather vague. But it sounds like you are running into a couple of problems.
First, you're doing something out of the ordinary. Or maybe the supplier is supplying something they don't normally supply. In either case, you've got design issues falling through the cracks that normally wouldn't. How you normally specify this is to say "Give us a Widgetizer Assembly like Company XYZ makes." But if no one normally does that, you don't have that option.
Secondly, it's not altogether clear from your post who should be responsible here. It's a structure "we" have designed, you say, yet other parties are now responsible for the structural design.

 

[enginerding]

Of course it is vague. We're on Eng-Tips.

A very good analogy would be a manufactured skylight being added to a building. We would have been the EOR for the building and provided design criteria for which the skylight had to meet. The other company was responsible for the design of the skylight to meet the design criteria provided (basically the 2006 IBC with wind and snow loads, etc. specified).

Rather than submitting a design for the skylight in one package, it seems the manufacturer employed 4 or 5 different engineers each to design one component of the skylight. We were given a pile of their individual calculations, but no complete set was provided that put everything together and accounted for global stability. The argument by the manufacturer is that they provided sealed drawings and calculations so they met their specifications. The powers that be allowed them to attempt to install their component, and thankfully (that it happened during installation instead of later on) it did not survive the attempt. Because of my objections all along, we are getting another opportunity to clarify what we are looking for - which is unchanged from before - 2006 IBC.

I think requiring one EOR for the component should make the difference.

 

[Dinosaur]

I still have trouble believing the cause of the original failure could not have been identified before materials were shipped to the jobsite. I am often required to review other engineers work for my agency, and I have sent stuff back for redesign all the time. It is a pain in the neck because I have to cite what didn't meet spec or code, but if a failure is that obvious, the cause has to be on the drawing somewhere.

I once had to send a post design back to the engineer because there was no LTB bracing. The maximum moments and shears were within the allowable but it needed to be braced to be able to attain those loads. The engineer argued with me and said that was not the intent of the AISC code. My engineer talked to an AISC guy that said we were totally correct. I didn't need to call AISC, because I knew we were correct. I wouldn't call a phycisist to see if gravity was expected to remain 32.2 ft/s^2 for the next fifty years either. Some things just are a fact and I don't need to verify them. Long story short, he eventually redesigned his posts but came in months late.

Good luck.

 

[enginerding]

Dinosaur,

For some reason the details of the failure are very sketchy by the time they get to me. I am pretty sure I know the reason, and I have been warning of it all along. What I get in response to my comments about global stability and bracing is silence from the company submitting the info and a phone call from our client saying that they tried to install it (without addressing my comments) and it failed.

 

[graybeach]

Some ideas (not that I have this problem solved myself):

-Require that all vendor calculations be independently checked.

-Require that all work be supervised by an experienced project engineer and that the resume be submitted with the proposal.

-Require that the project engineer reviews all submittals before submitting to you and certifies that he or she has done so in the transmittal letter.

-Require that a detailed work plan be submitted before the start of the design including an org chart and a description of the design process. Require interim submittals that have appropriate level of detail and coordination for the design stage. Reject interim submittals that are just a pile of old specs and old details dumped on new drawings.

-Ask that conceptual work plan be submitted with the proposal.

-Require a meeting with the vendor during the design phase and have them present the design and details to you in person.

-Require that a representative from the company be present during installation.

Good luck!

 

[csd72]

enginerding,

A brief summary of part of 'Aims of Structural Design' by IStructE:

Three main factors for a design:

Function - the structure must fulfil its intended function.
Economy - The structure must be of least cost for the given criteria.
Safety - the structure must be safe.

Sounds like they missed at least two of these criteria.