Significant Figures & and code interpretation 5

[Robvh1984]

I am a stickler to sound logic, e.g. significant figures, etc. However, My question arises from maybe a more philosophical standpoint (or poorly assumed to be). I see structural engineering as having 4 main components where variability may or will arise. 1.Loading 2.Engineering Analysis 3. Construction 4. Material Strength/general behavior.

All else equal, as engineers we have most control over the 1st and 2nd part of those 4 items as loadings may vary greatly with application (requiring our judgement) and analysis is literally performed by us. Therefore, as a young engineer, I prefer to use the code verbatim (i.e., if f'c=3000 I take it to mean that f'c has been measured to be precise to (1) significant figure(ignoring scatter of points). Then, for example, if I calculate the rupture strength (7.5f'c^.5) one calculates something on the order of 410.7919181288745850927273371006. I would say that the rupture strength is most nearly 400 psi, and I shall carry as many digits as I can from this intermediate calculation through my remaining calculations. I think my physics/chemistry professors would agree based on the standard mathematical laws or error propagation for arithmetic and single variable functions that this is valid.

However, My boss does not follow this logic when he performs calculations. He crudely & significantly rounds down whenever numbers dont seem to 'feel' right. Then in the end he selects sections with capacities as much as 20% greater than demanded by his rounded calculations. He justifies his methods of computing broadly based upon personal liability, loadings, and construction & material variability.

I find his methods to be unacceptable, as he applies this conservatism very broadly to everything he does. I understand our field is mostly experimental, but I feel that many of us (structural engineers) are so afraid of variability, we fly through analysis with grossly over estimated values and thus structural members.

In, short, how does everybody frame their logic in performing calculations as it relates to precision?



-Robert Miller, E.I.T.
KPA Structural Engineers

 

[Teguci]

I'd hate to be an engineer that doubted my numbers by 20%+ as a norm. Thats just lazy. I admit there are times where a quick calc is in order and a 20% "adjustment" meets the scheduling need of the client, but that should never be a standard practice.

The philosophical 20% is an admission that you don't fully understand the situation and you are attempting to quantify that fear.

Best response to your boss would be to ask why he doubts the numbers being calc'd.

As for precision, I don't reenter the numbers in my calculator to match precision and I rarely write down more than 3 significant digits.

 

[JoshPlumSE]

I experienced the same thing when I was a young engineer fresh outta school.

Once you start leading projects of your own (and stamping your own drawings), you'll have a much greater respect for making conservative changes based on what "feels right".

It also depends on your client / project. If your project is a "fast track" job that is likely to experience lots of changes when it gets time to build the thing, then thowing in some extra conservatism is a wonderful (and potentially inexpensive) way to save the project some major headaches down the line.

I remember working with one client who wanted every footing to be optimized to the most efficient size so that they could save money on concrete. Though it didn't help them any in the long run. In the end, the contractor basically threw away our calc packets and asked if they could just standardize on one or two common footing sizes. The client may have saved some money (because the extra concrete cost was absorbed by the contractor), but they partially compromised their schedule to do it.

 

[BAretired]

I agree with both Teguci and JoshPlum but there is another aspect to sig figs that might be worth considering.

In the calculation of a simple beam, the theory is well known, so the results should be as precise as the loading information. I would perform the calculation, then select a beam with adequate strength and deflection, but if there were fifty beams and three were a bit lighter, I would specify the same member throughout to avoid confusion.

In the calculation of the additional torsional strength provided by vertical stiffeners in a steel beam, I would neglect the added capacity because the theory is not clear to me.

BA

 

[msquared48]

GEE... To what thread are you referring BA?

I, too, in accordance with the KISS principle, add more material as warranted to help avoid the possibility of mistakes in the field. Relative to the mistakes, the cost of the additional material is minimal. Plus, it adds additional calacity to local areas of the strucgture where, if mistake is made, additional capacity is already in place so that a repair may not be required.

I never toe the line.

Mike McCann
MMC Engineering

 

[woodman88]

Given that in structures design the design loads are barely accurate to two significant figures. I feel that using a maximum of three significant figures is more than sufficient for the calculations.

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.

 

[msquared48]

I only use one significant figure - my wife.

Mike McCann
MMC Engineering

 

[Robvh1984]

Mike,

I feel you are right. If my wife says use a W12x19 and I do not do so, the personal consequences may be greater than the professional ones. HA!

-Robert Miller, E.I.T.
KPA Structural Engineers

 

[racookpe1978]

Well, I don't about your wife, but "my life" (er, wife) always makes me add just enough extra material so the final number cleverly rounds up to the next cheaper "standard size" that the vender can provide in time, but stay within the boss's boss's budget.

Fortunately, my boss usually agrees with her.

 

[Robvh1984]

Racookpe,

I do not disagree with you there. I do not think anybody would. But your decision is quantifiable and easily documented. What I was describing was haphazard calculations on projects that do not require "on the fly" decision making.

Of course, when u visit a structure with immediate bracing needs, you go conservative.




-Robert Miller, E.I.T.
KPA Structural Engineers

 

[SlideRuleEra]

rjm5062 - I agree with you... but only up to a certain point. As you state, arbitrary rounding of intermediate calculations is not a good idea. A level of precision should be maintained until the final (numerical) result is obtained, THEN engineering judgment (or common sense when a spouse, etc. is involved) is used on how to apply this mathematical answer.

Where I disagree with you is on the need to carry forward an unusually large number of significant digits during intermediate calcs. The reason being that unknown errors can "creep" in along the way. In your good example calc for concrete rupture strength here is a possibility for an unknown error:

1. Is f'c = 3000 psi accurate to one significant digit (3000 psi + or - 500 psi)?
Could be two significant digits: 3000 psi + or - 50 psi
Or three significant digits: 3000 psi + or - 5 psi
Or even four significant digits: 3000 psi + or - 0.5 psi

My point is that you cannot always make correct assumptions based only on the numbers presented, unless a tolerance is stated.

In your example equation, I would be more likely to question the precision and accuracy of the constant (7.5) and power (0.5).

For myself, I agree with woodman88 that three significant digits is all that is needed for intermediate routine calculations. In your example I would say the rupture strength is 411 psi. Of course I say THIS since three digits is the accepted limiting accuracy of a 10" slide rule.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[Robvh1984]

Sliderule,

I agree with you and I Like your description of the permutations with the compressive strengths. I inherently want to question those miscellaneous factors too. I guess my assumption is that the those empirical equations (when applied to the correct application) code, represent a minimum acceptable threshold deemed to be fit for design.

If you look at regression curves in the commentary and supplemental code/design books, they are never situated in the "median" of the scatter plots. They always seem to lie to the lower bound of the data. My guess is so that there is very little (-) component of (+/-) a value provided by the code and that the designer can spend more time worrying about his own assumptions and error. (Again, let me stress the proper application as so I do not sound like I ignore variances in code formulas.

From that, I frame my logic that assumes those minimums to be pseudo-exact measurements, and follow industry standard analysis methods and carry mathematical operations and their associated sig figs with them.

I must say I wouldnt argue for my point on KNOWN controversial materials or design methods, even when they are loosely specified in the code books.

-Robert Miller, E.I.T.
KPA Structural Engineers

 

[Ron]

Back to my old adage....

We tend to measure with a micrometer
Mark it with a crayon
and cut it with an axe.

That's the process from design through construction.

Significant figures mean very little in the whole scheme of things, particularly with building construction.

Why is it necessary to compute stresses to two decimal places when the wind loads you used are up to interpretation in several respects, each producing a significant difference of stress levels?

Make what you decide follow reasonable logic, common sense and a valid standard of care.

 

[IDS]

I'd hate to be an engineer that doubted my numbers by 20%+ as a norm. Thats just lazy. I admit there are times where a quick calc is in order and a 20% "adjustment" meets the scheduling need of the client, but that should never be a standard practice.

The philosophical 20% is an admission that you don't fully understand the situation and you are attempting to quantify that fear.

It depends on the material. Some material parameters (including concrete tensile strength) have enormous variability, much more than 20%.

All engineers should be aware that they don't have exact values for material properties, and that code values are not always conservative, and should consider the consequences of material strength and stiffness parameters being at realistic lower or upper limits. This range will often be much more than +- 20%

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[gwynn]

The question on conservatism depends on what you are designing and who it is for as well. There are some contractors I work with who are consistently out of tolerance. If I am performing design work for one of them I will add in extra capacity. A few extra pounds of steel costs far less than tearing it down and starting over or trying to install remedial measures to bring capacities up.

 

[OHIOMatt]

Three comments

1. I would remove the name of your firm and your name from your tag line. We all now know that KPA structural engineers does not provide the most effcient, cost effective designs (may be valuable information to your competetors).

2. In today's world, we are all working under increasing pressure regarding price and timing. We have all experienced increases in competition from both a fee and timing aspect. If your boss feels that he can meet the budget and provide a safe structure, he make take some short cuts. Typically a 20% increase in capacity does not mean a 20% increase in cost and weight.

3. Codes are minimum standards of design. There are several areas that I routinely feel that the code is less than adequate.

4. If you feel that engineers have control over loading, then you better think again. We have control over how we model loading, but that often is not how loading is applied. I was just in a structure that by code and previous use had a floor live load of 500 psf. My obeservations were that a significant area of the floor was loaded to close to 1000 psf.


I guess in summery, engineering is not all science. There are significant buisiness and art components as well.

 

[MJB315]

Rob,

I agree- it may be best to take down your name and firm from your posts. Precision, significant figures, efficient design or not- all you need is your supervisor to see that you're categorically...characterizing, his design approach on a Monday afternoon and it'll become an aforementioned personal issue for everyone.

But like all things, I think it depends. Whereas deisgning a simple structure is mainly science-- putting together a large structure is an art.

Being able to put together a large drawing set is: 3 Parts Communication, 2 Parts Drafting, and 1 Part actual analysis and design. Figuring out whether the transfer beams should be W18s or W24s is rarely as time consuming as figuring out whether either works with the architecture- and whether it can get shown on the drawings in time. Sometimes, for the project's (read: schedule's) sake-- you just need to pick the W24 and keep moving.

Further, "feels right" is important. In fact, its what our entire system depends upon. At the heart of it, the only day-to-day safeguard our structures have is the fact one person ultimately needs to stamp it. There's not always a peer review; the code department doesn't look at the drawings with a fine toothed comb. So that one person, above all, has to feel right doing it. That's it. If not, over the course of a career, he'll never be able to sleep a day in his life.

The science just helps him do that.

-MJB

 

[IRstuff]

Unless you're are doing your calculations with pen and pencil and slide rule, the carrying of sigfigs is completely transparent tot he end result. On any spreadsheet or math package, truncating the precision in intermediate results requires substantial work to make it happen. The final result can, and should, be displayed at the desired precision, consistent with the data and the theory.

The fact that all your intermediate calculations are carried out with 17-digit precision should be irrelevant.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[handex]

"We tend to measure with a micrometer
Mark it with a crayon
and cut it with an axe.

That's the process from design through construction.

Significant figures mean very little in the whole scheme of things, particularly with building construction.

Why is it necessary to compute stresses to two decimal places when the wind loads you used are up to interpretation in several respects, each producing a significant difference of stress levels?

Make what you decide follow reasonable logic, common sense and a valid standard of care. "

Extremely well put, could not agree more

 

[SLTA]

The great quote:

"Structural Engineering is the Art of moulding materials we do not wholly understand into shapes we cannot precisely analyse, so as to withstand forces we cannot really assess, in such a way that the community at large has no reason to suspect the extent of our ignorance."

The 20% being discussed may actually be 12%, or 57%. Big picture - how much do we really understand?