Units, formulas, anyone? 4

[SKIAK]

So... this turned out to be more of a rant than I expected... sorry.

I check a lot of submittals. Most submittals are rebar shop drawings, concrete mix designs, that kind of thing. Sometimes I get submittals for cold formed steel curtain walls designed by others and things like that, and they have calculations attached. More often than not the calculations are very vague, employ some type of software printout that is completely foreign to me, and any hand calcs are not accompanied by formulas or units. In my designs I have to do all of my calculations by hand. Software can assist me in choosing the best shape, but I cannot turn that in for a design. Maybe I'm jealous?

When I was in school (less than a year ago) there were some very basic principles that needed to be followed, such as: No units? No credit. If we were allowed to use something like the AISC Steel Manual and we used formulas from it, No reference? No credit. My professors always told me “if I can't look and see what you did it’s useless. People in the industry won't stand for incomplete calculations.”

Getting to the point, I have reviewed so many calcs that have only solutions with no work, some work but no labels or units, or (my favorite) "[beam, column, footing, ect.] OK by inspection." Is it wrong for me to see this phenomenon as any of several things: sloppy, lazy, unprofessional, and/or arrogant? I think it’s great that some people are so good at what they do they don't actually need to verify their design with numbers. Wonderful. In all honesty, I AM jealous, but I'm expected to approve this? How can I possibly check and approve anything when there isn't actually anything worth something to look at? Most often from my supervisors I get "I'm sure they know what they're doing, they've been working there forever, it looks OK." OK, let’s assume they have the competency to do that kind of thing, but are we to assume that they didn't make a mistake? Isn't that what checking is all about? How do you check off on "OK by inspection?"

Structural Review Comments: Pretty pictures, straight lines, good handwriting. Calcs? Eh… ya sure, he/she is probably right.

I understand that some calcs are very basic but is it too much effort to write out a formula? Use units? Labels? Give actual insight into design methodology? Isn’t this Engineering 101?

Is this common or am I being overly sensitive?

 

[SylvestreW]

I can appreciate standard designs, and as long as it looks like there is an understanding of the conditions I don't say anything. I have only once returned a submittal because my check on their design failed. Turned out that they were using a standard template and didn't update it for the wind zone the building was in. Other than that I try to run my own numbers if I can't use theirs, if they come out similar I don't say anything. It would just be quicker and easier if we didn't both run through all the numbers. I'll save my demands for when I'm a little more established.

In a way, you are likely learning more and becoming better at your job doing it this way than if you did have the numbers fully and weren't forced to dig.

-
Syl.

 

[apsix]

I can appreciate that the response to this thread from an (Aerospace) would differ from that from a (Structural).

 

[monkeydog]

True, in Aerospace, you have one chance to get it right.

 

[zelgar]

As a government employee (State), there has been many times when I have requested additional information from a professional engineer over his calculations that did not include formulae, units, assumptions, sources of data.
I am not saying that the calculations are wrong, but did not contain sufficient amount of information for my review. Sometimes I'll be able to halfway follow what was being performed, and the request is to verify my assumptions.

Any engineer who is not willing to provide the basic information for calculations (e.g., formulae, units, sources) is one who is wasting the time of others who need to review their work. And he better pray that something critical is never missed due to his negligance (e.g., resulting in deaths or million $$ project).

 

[SKIAK]

I guess thats what gets me the most. I don't really mind doing the checks or running the numbers (in fact, its kinda fun, like a jigsaw puzzle with problem solving), it just seems to me that a design without usable calculations isn't complete.

For example, this may be an extreme case? A building owner came to us, when he was trying to sell two warehouse type buildings, for a structural evaluation. Pretty basic pre-engineered steel buildings (approx 60'x120'), and a local engineer had done the foundation. The buildings had been up for around two years. I went out to the site, took some pictures, looked around, everything looked good in building #1. Building #2 was identical except he had had a two story timber framed office built inside at one end (approx 20'x60'). Apparently the contractor did the design and fabrication but I got rough layouts/dimensions/materials knowing already that this was going to be a problem. Got back to the office my boss told me to check the foundation calcs and do a seismic evaluation of the inside office structure since we had no previous calcs for that. The previous engineer that had designed the foundation was vague in his calculations (approx 4 hand written sheets with diagrams for the entire foundation). The loads were given by steel building manufacturer but still needed to be combined (as was explicitly stated in the building manufacturer's drawings). As best I can tell the foundation engineer did not combine the loads and just picked the biggest one. He had drawn very nice moment diagrams for the grade beams, but how were they calculated? No idea, mine were much larger. In fact everything I calced was on the order of 1.5-2.0 times larger than the previous engineer (moments/shears/bearings). At the start my boss told me that this guy has been around forever and he wasn't worried. I tried to do all the back calculating I could figure out but came nowhere close to what I felt like was a legit design. My guess is he has done it a million times, second nature and basic to him. The "shearwalls" in the office building on the short sides had less than 10' to develop the shear resistance, 1/2" gyp board nailed to cut 2x4's (cut to fit around and inside the channels holding up the exterior siding) was inadequate (to say the least) of carrying anticipated seismic forces in a place where seismic design can often control (Alaska). My boss had me write a report that detailed out the deficiencies I found with my calcs (approx 20 handwritten sheets without drawings), don't know what happened with it beyond that. I can only hope that I was wrong (and that somebody checked MY calculations!).

 

[whyun]

An engineering package shall be self-contained and in itself, fully describe the engineer's design. It is not acceptable if a "competent" reviewer has to physically communicate with the engineer to understand what he is doing. At the same time, it is not the rold of an engineer to "train" the reviewer.

 

[GregLocock]

"True, in Aerospace, you have one chance to get it right. "

I'd argue the opposite. In aerospace and automotive we have comprehensive test programs so that any initial errors by the engineer are caught before Job 1.

These structural guys don't have test programs, and the prototype /is/ Job 1. That's why design to code is such a strange paradigm to us, and optimal design for function is such a strange paradigm for them.



Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[monkeydog]

True, once the baseline design is qualified, you should have screened any engineering errors. But, engineering modifications to the design does not necessarly result in requalification of the design. Also the margins to failure could result in catatrophic failure with the exception of manned programs.

 

[JStephen]

I have mixed feelings about this whole issue.

First off, it seems that the purpose of the calculations is to do the design. The calculations are not intended to educate those unfamiliar with the topic. You should not request calculations and expect a textbook.

What is the purpose of submitting calculations for review? As best I can tell, in most cases, it is just a "feel good" thing to see that the calculations exist. The ones requesting the review have no idea how the calculations ought to work. You can quote equations and plug numbers in them, but you could quote different equations and get different numbers and nobody would notice the difference, just so the numbers match the equations.

If there is to be a technical review for accurary in the calculations, it needs to be by someone thoroughly familiar with the topic, not a matter of handing them over to "the new guy". Unfortunately, this seems seldom to be the case. Professional competency is critical when preparing the calculations, but is thrown out the window when reviewing them.

Through experience, I have learned that the more detailed and complete the calculations are, the more likely you are to have criticisms of minor issues and have to go back and forth on that. So the best strategy from a submittal standpoint is to just omit all the minor items from the calculations, give the reviewer a few major items, and everyone's happy. The one exception to this rule is when you have extensive spreadsheets or computational analysis results, which nobody is expected to check anyway. But don't assume that the calculations you review are all that were done for the job- that may or may not be the case.

Additionally, part of engineering judgment is knowing what to calculate and what not to calculate. Anything that can be designed, can be designed to any desired degree of complexity. If NASA designed houses, they would have a dozen engineers spending a year on each board. When the work requires an experienced engineer, and that experienced engineer says item X is okay by inspection, you are getting part of that experience that you payed for.

 

[HamishMcTavish]

The bottom line is that if you are approving something then do not put your name to it unless you are satisfied that the calculation stands up to scrutiny. No engineer should be expected to rubber stamp someone else's work - it's unprofessional.

Would you rather ask pokey questions (and face a little heat for doing so) or be being asked them by a lawyer when something has gone pear-shaped and the buck stops with you?

No more things should be presumed to exist than are absolutely necessary - William of Occam

 

[whyun]

I agree with JStephen. Review of anything, ideally, should be done by someone of equal or better caliber. Of course, it is very subjective. As a minimum, PE's work should be reviewed by a PE and SE's work should be reviewed by an SE, which isn't the case in many jurisdictions.

Submitter should not prepare the calculation package in an effort to educate the reviewer.

 

[dik]

When I've been asked to review a design, I usually check to see that all the info for construction/or whatever is presented and add procedures as required. I then do a preliminary design using a methodology that I'm familiar with to see that the values are in the same ball park. It is very seldom that I check items by reviewing the actual calculations prepared by someone else, except if I encounter something that doesn't seem to fly and I'm trying to find out the reason for the discreptancy.

Dik

 

[Dixon80]

SKIAK - ask for what ever you need. I've been a state regulator for about a 1/4 of a century (that puts some labels on things). Starting out in the Water and Wastewater fields, now just in Water. Even with a preliminary conference in which we tell the consultants what we expect in the way of submittals, I'd say only 10% to 15% are approvable on the first submittal!
Yes, I started out reviewing more experienced engineer's work, but not only was I requesting changes; but at that time my work was checked by a PE. Now that I've had my PE for two decades, my work is still checked by another PE.
Dixon

 

[cry22]

Greglocock!
While I agree that your statement above is accurate, I have to say that the "optimal design for function" theory is what gets those cars today totaled at 5 MPH, compared to when they made them solid (i.e. to code).
In 1993, the first Trade center bombing did not bring the towers down, thanks to the "design to code" and not "for function".

 

[Dinosaur]

I also believe the varied opinions are indicative of industry expectations. Even in structural engineering, there is a big difference between the building industry and the bridge industry. Also, there are different expectations for turn-key products vs. bid designs built to plan.

For my part, I have designed many things, and I have reviewed the designs for many things. I look it over, and tell my boss what I think I will need. He either buys in or he changes my opinion. Most of my bosses would rather let me handle it my way than get involved.

I often will spot check a few numbers along the way. If I can't get anything from the calcs, then I'll ask specific questions to get clarification. If they can't answer my questions, then I have to question if they can even do a design.

I have caught some catastrophic errors in the past. I never take that job lightly.