Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

RE: Proposed Standard of Care "again"

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Paul Crocker: Thank you for your comments. Much of what you've said is true
- but I don't feel that changes the benefits of what I've proposed. Below
are some responses to your comments. 

> You seem here to equate "unchecked" with erroneous.  

I believe that it is frequently true that "unchecked" equates to erroneous.
I have been involved with checking of structural calculations for over 25
years and rarely find a calculation that does not have errors. Errors may
include math errors, erroneous assumptions, misapplication of design
criteria or code requirements, design omissions, misapplication of design
methods, etc. More experienced engineers tend to make fewer mistakes but no
one is immune to mistakes. It is true that not all calculation errors result
in significant changes to the resulting structural details but often design
changes are required as a result of such errors. 

> Many of the notable structural failures throughout history were checked by
a 
> number of different engineers, each of whom failed to see the problem for
a 
> variety of reasons.

I too have seen designs which have been "checked" which still have inherent
errors. I feel we can  relate engineering checking to field inspection of
construction. Little or no inspection/checking will increase the probability
of defects getting thru; frequent inspection/checking significantly reduces
the number of defects which get thru; but nothing is 100% effective. Others
have recently pointed out that Structural Engineers are responsible for the
safety of their designs to protect "thousands of lives". Having defined
checking requirements is a small price to pay for this protection. 

> In that light, I'm not sure on what basis you assume that the designs you
> reviewed were "unchecked" unless the engineer's design methodology came
out
> in a later court case.  Perhaps the engineer who checked them also missed
> the error.  

Yes, I have often "assumed" that submitted calculations were unchecked. This
assumption was based on the fact that there were no checker's marks on the
calculations, if there was a space for a checker's initials on the calc
sheets it was not filled in, and there were readily apparent errors in the
calculation which affected the results. If the calculation was checked, the
checker did not do a good job of it. 

> In some firms, checking "flows downhill" and thus each level that it
passes 
> through is less capable of finding problems than the one that they
received 
> it from. "Checked" is a very broad term in that sense. 

I concur that the "quality of checking" can vary according to interpretation
and application. But requiring some level of checking is at lease a start. 

> Like anything else, the reputation and resume of the firm tell more about
> their work than anything else.  Even if a specific regiment of checking
was
> mandatory, you would still see wide differences in the standard of service
> offered by firms based on staff experience, staff level of education,
level
> of continuing education, workload and overtime, software used and
experience
> with it, office conditions, and even general temperment.  Checking is
> crucial, but it is a long way from assuring a uniform level of service
from
> different firms.  

All of this is true. Quality of construction can also vary depending upon
the contractor one hires - but field inspection is still a good idea whether
you have a "good" contractor or a "bad" one. Even within one engineering
firm, quality of design can vary by the individual. 

In the case of contractor submitted designs, I don't have a choice in "who"
the contractor uses to do the design. So again I would ask: Can I require
that the contractor's engineer have their calculations checked by a
qualified person other than the preparer if it is not explicitly in the
written contract? What "standard" can I point to when I receive an
"apparently" unchecked design?