sharpening your pencil

huangyhg

sharpening your pencil
how much do you "sharpen your pencil" when doing structural analysis and design? in other words, how conservative are you when making assumptions?
we have all made assumptions that simplify our calculations, as many times it is impossible to model actual conditions. in my experience, i have noticed that younger engineers try to account for every little detail and tend to over-design. the flip side to this are engineers who make frivolous assumptions or over simplify almost to the point of negligence. i suppose we must try to fall somewhere in between and keep the client happy at the same time. i've often heard the phrase of an engineer "sleeping well at night" meaning that he or she is not worried how safe a particular design is. do you sleep well?
anybody have any thoughts?
motorcity,
we geotechs have a similar tightrope to walk - if you're too conservative, your recommendations are too costly and  the client won't hire you again; if you're too aggressive, you pray that the structure and the soil are as good as you hope they are and eat tums like they're candy.
i personally tend towards the conservative side of the street and then try to look for places where i have a little cushion and try and sharpen it up a bit.
it helps to run things by your colleagues who can point out where you can get aggressive without going overboard.
good luck!
jeff
motorcity,
you use the term "keep the client happy."  many times the client wants you to tell him everything is ok just like he has been doing it for years.  that would make him happy.  sometimes he would like you to stamp his drawings without doing any calculations.  that would really make him happy.  it seems that all the responsibility is put on the engineer and very little on the customer who is happy to take the lowest bidder.  
regards,
-mike  
i tend towards conservative assumptions, unless i know that places where i can really sharpen my pencil reflect real life closely.  the important thing is to know what assumptions are used and how close to real life they are.  the more uncertainty, the more conservative i tend to be.
these are scenarios when i'm forced to make assumptions and be conservative:  a project schedule that's really tight such as design build, the architect and other consultants are not going to be finished before i need to provide foundation and framing designs; when working on renovation/addition to an existing building where drawings of that building are not available and i don't have visual access to the existing structural systems.  otherwise i try to "sharpen the pencil", gather as much info as possible and design more economically.  of course, as far as my clients know, everything i do is the most economical design possible.
a follow up question:  how conservative is your fee proposal when the only information you have is a verbal description of the building?  do you add alot of "fat" in the fee to cover unknowns?  for me, when a client asks for a fee on a single story, 20000 sq ft building, steel framed and that's all i know about the job, my fee can be as much as 25% higher (depending on experience with the client) than if they give me a preliminary plan, elevations and building section for the same building.
i am as conservative as my resources require.  if i have time, i save the client money.  if i don't, i can't.
in design of structures, i tend to be conservative in estimate of dead loads to allow for future modifications but more accurate in analysis and design.
i also tend to be more conservative in connection design than   
my approach has always been virtually identical with whyun's. additionally i try to look at differential construction costs. an example would be whether to make an industrial slab either an "acceptable" 10 inches thick or a "conservative" 12 inches thick:
1. subgrade preparation identical.
2. forming virtually identical.
3. rebar materials & labor similar.
4. concrete placement, finishing & curing similar.
result: differental cost for 12" slab is essentially the purchase price of the extra concrete. go with the 12 inch thickness.
what does "sharpen your pencil" mean? i don't understand the reference.
tomfh-
"sharpen your pencil" means to do a more precise calculation.  for instance, when estimating your dead loads, you can use numbers that you know are conservative through experience, or you can "sharpen your pencil" and add up each component to the second decimal place, thereby reducing your calculated dead load, but increasing your design time.
back in the old days, these calculations were done with pencil and paper.  supposedly, you would want to sharpen your pencil before carrying out a precise calculation. to save paper, i guess.
i tend to not be too conservative.  if i have a fairly straightforward calculation, i use the most economical size.
the reality is that our designs are conservative even without our additional help.  the factors first of all of course account for a degree of that.  something a lot of people overlook is that the design loading is often much much higher than the actual loading.  for example, an office building is designed for 50 psf live load and 20 psf partition load, or 70 psf total.  a study was done where they went into many different types of buildings, and weighed everything there to come up with the actual loading.  some of the results they came up with are:
live load:
11 psf for paper office
8 psf for electronic office
6 psf for residence
4.5 psf for hotels
12 psf for schools
-this information can be found in the appendix of asce 7 (table c4-2, page 260 of asce7-02).  these are averages, and so some areas in the office, for example, that you design will see the 50-70 psf, but in general, your design is very conservative.
as for sharpening my pencil, i always find the most economical section for the major items such as joists, beams, slab (if applicable) and pier sizes.  i almost never check connections, or typical details such as bracing sizes, lintel sizes, roof opening framing, masonry reinforcing, etc.