how much do you keep in the way of calcs

huangyhg

how much do you keep in the way of calcs
i recently got to run my own project for the first time.  it was a very small project, but i was the one coordinating with the architect and doing all of the detailing/design.  
this is a question for the experienced engineers out there.  how much of the calculations that changed during design do you keep for the project?
just as an example, i designed some clips to support a 17' strip window 6" from face of wall.  initially these clips were carbon steel (fy=36 ksi) with a width of 3" (for architectural considerations) and i designed the thickness to be 1/4".  
then the architect changed it from carbon steel to stainless steel (fy=30 ksi) and it needed to be 4" wide with 1/4" thickness.
then i talked to the glass guy to get an idea of how much bearing width they want to see (turned out to be a minimum of 4") and found out that they don't like to use ss greater than 3/16" so they can maintain a tight radius on the bends (it was a bent plate).
the final product ended up being a 3/16" thick x 5" wide bent plate.  after all of that, i have three different designs for this one simple bent plate.  
i am sure others have been down this road with similarly small things (and i am sure with larger things).
the question is whehter there is any reason for me to keep the first 2 designs for the bent plate with my organized project calculations or if they can just be tossed out.
any insight here would be greatly appreciated.
check out our whitepaper library.
i take out what does not apply to the building anymore or scratch void through it. sometimes i write over the original when it changes in order to save time instead of completely re-writing a calculation. i can't stand to do something more than one time and this goes for everything in my life not just at my job, so i tend to wait until later to get a lot of my calculations together and get the drawings to around 80% first. this is because i know if i do something early it is going to be changed a couple of times before the end just like what you went through.  some things i have to do early just to get the project going and for dd estimates. spreadsheets are the best since changing numbers takes seconds.
i think you should keep your previous calculations and write 'void' or 'revised' on them, this way you have evidence of the architect changing things if you are ever questioned as to how you spent your time on a project. if using spreadsheets, keep a note in the file or write 'void' on the previous printouts for record.
not a structural here (geotech) but i do run across similar situations when foundations/loads change. i might trash the first design runs when i purge the file, but if the design changed/evolved due to new/other information, wants, needs, etc then i'll make notes on the final copy so that i can see how i ended up with my last round of assumptions and design parameters. this is due to in case something goes to litigation and the lawyers ask why you used "this" when the original documents said to use "that". it just seems like a good idea to keep track of how the design evolved based on project/owner needs/wants. i wouldn't think that it's completely necessary to keep every single calc for a design that was not used...as a matter of fact, it seems like more of a liability to me...what happens if a lawyer finds an error in a design that was for the project but was not used? would that be turned on you some how? (thinking outloud there--i don't know the answer). and everyone else doesn't necessarily keep accurate documents...especially when it comes down to you having to be the one on the hook for something. that's my opinion for what it's worth.
keep the the other versions until the project closes out.  the contractor may have different ideas during construction.   after that, you clean out your calc file and keep only what you absolutely need.
as stated above, a detailed calc package can be used against you in some cases, so just keep that in mind.  especially if you did any quick and dirty approximations.
we keep everything.  our copy machine scans to pdf so we archive even the ugliest "chicken-scratch" calcs.  we also just save revisions of spreadsheets or risa files.  our opinion is that hard drive storage is so cheap, why delete anything.  at the end of a project, we organize the calcs such that the final design calcs are easy to find, and the other junk goes in a subdirectory.
i keep a copy of the nice, numbered, headers-filled-out sheets in the front of the calculation volume.  i then have a tab and insert the equipment sheets and other external back up information. behind that's another tab with all my preliminary calculations, sketches and miscellaneous notes and information.
long story short, i keep everything, mostly to provide a recollection on how the design evolved and in case anyone ever asks me to justify my hours.
keep all revisions up to the final design, put in a notebook with good table of contents, require your company to keep in archives for a minimum of three years.
i would say the easiest way to handle miscellaneous calcs that you think you don't need them any more is to scan them as pdfs and dump them in the project folder as miscellaneous calcs. this way you can keep your calcs clean, avoid confusion and at the same time you have a chance to print your old calcs if at all you need them, which i think is never gonna happen.
strucguy
i cross out the part of the calc that is revised due to changes.  in my opinion, it's all part of the work product and thought processes leading to the drawings we seal and sign.
several years ago i spoke with an engineer who indicated that his lawyer's advice was to destroy all his calculations after the project was built.
the lawyer indicated the first thing a law firm would do if they filed suit on one of his jobs, would be to hire another engineer to go over all the calculations.  the laywer's advice was not to give them information that they could use in a law suit.