Calculation Peer Reviews 3

[skeletron]

I wanted to start a separate topic for my post in thread507-459691 to see if there was any further discussion.

With the discussion trending towards the tools and processes for creating calculation records, I would be interested in understanding the backwards discussion:

What are your tools and processes for checking calculations?
What are your preferences for calculation submittals?

 

[Lomarandil]

Tool and Process: when checking calculations, I ask my junior engineers for PDF output from Smath (or MathCAD). I then use bluebeam to make comments and markups.

I strongly believe that calculation submittals should tell the story of the engineering decisions and assumptions. Printing FEA output is not a calculation, that's record keeping. Submittals should include sketches, notes, even a paragraph or two of narrative. Include spreadsheet or FEA output when needed, and highlight or mark up the important inputs/assumptions/outputs.


----
just call me Lo.

 

[Aesur]

I second Lomarandil, although I don't require a narrative unless something needs explaining.

 

[Agent666]

Trust me the narrative helps in 10 years time. It's easy doing the peer review and coming up against some confusion, as you can always ask the designer at the time to clarify. Often this is for your own benefit/understanding and they wouldn't go back and amend the calculations on record.

In 10 years time you don't have this luxury, you're left with numerical gibberish. Trust me sometimes you cannot even tell what people are even checking, maybe its obvious if you've been balls deep in the job for a year, but as the reviewer looking in from the outside, if you ain't explaining things you ain't doing design right.

I was always taught the calculations should tell a story, write as much as you need to get the point across. Asking the reader to make leaps of understanding or interpret numerical gibberish is not engineering.

A wise engineer once told me the actual calculations are almost secondary to clearly outlining things like loadpath, design philosophy, etc. If the load path is clear any engineer worth his salt should be able to prove it works. If the load paths are dubious and the calculations don't explain the 'why' then your life as a reviewer gets hard. If you step back to reviewing it independently often you simply cannot justify what has been done using good engineering judgement. Peer reviewers are not mind readers, not all engineers were created equal.

I feel these days these things, being the essence of good engineering are lost on less experienced engineers, in the sense that at the first sign of trouble they create some elaborate computer based calculation without really explaining (using words) what they are even doing sometimes.

 

[Aesur]

Agent666, I understand what you are saying and agree they help, but also have come across in my region that many times calculations are thrown out after only a few years and are rarely available on projects older than 10 years. I believe the main reason in my area that the standard of care doesn't include these narratives is simply because it was one of the last places that time could be cut to meet the small fees expected, other than paying engineers less than $20/hr. In my region medium to large size firms are hurt by the 20+ one man shops in town that will do some of these projects for less than $0.25/sqft (luckily many are getting close to retirement).

 

[Agent666]

In my region in some areas, the calculations are recorded on a property file with the local authority for all eternity. Others you almost don't even need to submit them.

I've seen projects calculations with little narrative, engineer leaves, something needs to change, might as well start again instead of a few minutes to follow through the same logic for a different member size for example. It's like a builder walking away from a job when it's half finished and saying he finished in my mind. Maybe my own personal standards are too high, but I've seen my fair share of bad, brief, calculations and more often than not brief = mistakes because not enough thought, analysis, time has been put into the calculations or the documentation.

 

[canwesteng]

A paragraph of narrative takes all of a few minutes, it's sloppy not to include one.

 

[Aesur]

I am curious what you include in your narratives that isn't included in the key plans? Every calc package I do has a key plan for every beam, column, lintel, joist, etc.; it's fairly self explanatory. Same goes for the lateral system, indicating all shear walls, how the loads were applied and so forth. Just not sure I agree that a narrative makes that much of a difference for properly laid out calculation package. Now if you are using RAM or some other project that just prints out loads of data, then sure you should consider explaining some things, but I don't see a point in explaining how I calculated a beam, how I came up with the loads on the beam, etc.. things that are in the actual calculation or load sheet.

 

[skeletron]

I agree with including narrative/comments in the calculation package. They help explain how and why a decision was made. I think it also makes the calculator accountable for acknowledging when logic is used to proceed.