Eng. practice of allowing 5% overstress 33

[radair]

It seems to be standard structural engineering practice to allow up to 5 percent overstress in structural design. It's been this way since I graduated college in 1980 and I've seen the practice commonly used in the tower analysis field for the last 15 years.

I've been asked my opinion by a government agency as to why this is a safe and acceptable engineering practice, including citing any relevant structural codes. They are not questioning my work but are asking me for a signed & sealed letter of opinion. It seems to me that this would be a better question for their state engineering board of licensure.

Can any of you help?

 

[hokie66]

But that is the point I was trying to make. The bridge did not collapse because the engineer decided to accept a slight overstress in a particular member or connection. It collapsed because a gross error was made somewhere along the way in the preparation of the drawings.

 

[frv]

hokie66-

Exactly!

Didn't they find out that a gusset plate (a critical component in a statically determinate bridge) was half the size it was supposed to be? And it still took thirty-some odd years for the bridge to collapse! And remember that for years the bridge was also seeing much higher traffic loads than the original designers had anticipated.

And people are worried about 5%?

 

[civeng80]

Remember that the people who judge are not engineers but legal practitioners. They have limited time and can be careless and downright irrespossible in making their decisions in my opinion. But it is their job. I beleive that calculations revealing an overstress would be a minefield for the prosecuting lawyer.

I have been in a 6 month court case (not engineering related) and it was pure hell.

 

[JStephen]

Where I have run accoss this before was at a former employer, where they routinely sized foundations for 105% of whatever soil bearing capacity was given in the soils report. As justification for this, I was shown some old textbook that said 5% difference in the bearing values was negligible.

It seems odd to me to say, on the one hand, that this 5% is so trivial that it will never make a difference. And on the other hand, to say that it is so vitally important that we must do it to stay in business. It can't very well be both.

It seems to me that most of the arguments that you could make for allowing a 3% overstress across-the-board could also be used to justify a 5%, 10%, 15% or 25% overstress as well- sort of a slippery slope once you get on it.

 

[frv]

JStephen-

There is a difference between 5 and 25%. Where you draw the line is a matter of engineering judgment. I feel like I could defend myself in court for allowing a 5% over-stress. Don't know about 25%.

 

[strguy11]

I had this question come up on a steel building, and justified a 3% overstress based on material strengths. Even if you are specifying Fy=50 ksi, the steel you get is probably closer to 58 ksi. (At least this is what I have typically seen on mill certificates). The reviewer was reasonable, and accepted this answer.

You could use a similar argument for concrete, as concrete strengths will always be higher.

For other materials, I think most allowable stregth/stress values have been determined by testing, and safety factors are applied to the values given, so a small overstress is ok.

I typically limit it to 3%, but 5% would be an engnineering judgement.

 

[csd72]

frv,

Where you should draw the line is specified by the code!

 

[miecz]

It matters whether I'm looking a a single element or an element that is repeated throughout the structure. For a single element, the cost of bumping the size will have insignificant effect on the total cost of the structure. For a member that is repeated throughout the structure, it justifies going back and fine tuning the loading, boundary condition, or other assumptions to get the strength within code, rather than allow a 3 or 5% understrength. In either case, you either meet the code, or you don't. If you would feel comfortable standing up in court and saying "in this case, my engineering judgement overrides the judgement of the people who did the research to write the code," then you're a better man/woman than me.

 

[frv]

miescz-

Nearly every professor I've had (in engineering anyway) is directly involved in code writing and/or research relating to the codes.

Invariably, they all stressed using our judgment when it came to things like this. If you understand the statistical nature of loads and resistances, you understand the magnitude of 5% (please see my post on DL and LL for metal roof deck), understand that materials will have a significantly higher resistance than specified (much more than 5%), that your assumptions are conservative to begin with, then you can justify, IF YOU SO FEEL INCLINED, a 5% increase in stress without having to go back and revise your calcs. If ever it came back to court, you could then annotate your calcs to prove that whatever went wrong, it was not your fault.

So no, my judgment doesn't supersede that of the code writers. It understands the intent of the code writers.

And no, I'm not a better man/woman than you.

 

[civilperson]

I defy any court or reviewing body to find a 3% over stress discrepancy in my structures since I will not produce the raw calculations for their use. When they must produce the loads, then the model and then the analysis, many decisions as to what to choose as relevant become involved and agreement within 10% of my calculated stresses in the various members become problematic. When I review other's designs, any answer that produces less than 10% overstress is OK since I obviously used different model and assumptions as to loads/load path. Only buildings with prescriptive code requirements can be compared to design analysis and have reviewer say that the particular member is overstressed. Most codes have provisions to lower total loads on columns and footings due to the probability of all floors receiving the maximum code loading. A good design with no mistakes can be overloaded and made to fail by a unique cicumstance, (i.e. a 747 crashing into the structure with 12000 pounds of fuel).

 

[miecz]

frv-

I've been away from school for so long, I haven't spoken to any code writers for years. It what you write is the general thinking, then it should be easy to get someone to write it down. If I were the original poster, I would simply write to AISC or ACI and request such a statement.

civilperson-

We're required to submit our calculations with the final plans.

 

[radair]

Thank you all for your thoughtful input.

This instance involves analysis of existing telecommunication towers regarding their capability of supporting additional antennas. I have a number of examples of other engineers using a 105% maximum.

I am going to decline providing them with a signed and sealed letter, as this question does not involve a project of mine. I will likely give them a response by email and incorporate some of your comments.

Thanks for the help.

 

[Teguci]

IBC 2003 states under chapter 34 - 3403.2 - Additions and alterations ... not increase the force ... more than 5%.

synopsis - Code allows a 5% overstress for existing structures. For new construction, this is a no-no except for as-built pile cap designs.

 

[tngolfer]

Look at it this way.

The IBC has decreased the LL in garages 20% (50 psf to 40 psf) since the 2000 edition but it is considering going back up to 50 psf. Does this mean they are going to close all of the garages that were built in this time frame? No way owners and lawyers would allow that.

Does that mean that the garages will be 25% overstressed? Probably not.

There is so much "play" in the loads and actual stresses that who knows what the actual stress in the member is and if it actually 5% overstressed. That being said, to design per the code you have to keep the ratio less than 1.0.

 

[radair]

Teguchi wrote: "IBC 2003 states under chapter 34 - 3403.2 - Additions and alterations ... not increase the force ... more than 5%.

synopsis - Code allows a 5% overstress for existing structures. For new construction, this is a no-no except for as-built pile cap designs."

No, I don't believe this can be interpreted this way at all. The Code states that "...additions or alterations to an existing structure shall not increase the force in any structural element by more than 5%". This is not addressing overstress, but forces on structural elements. Unless your members were designed to 100% of their allowable stresses this does not apply to my intial overstress question.

 

[civilperson]

meicz,
Who mandates the calculations are submitted with the plans? If I state OK by inspection or model with a simplified analysis, (i.e. simple hinge connection with shear tab between beam and column or pinned base plate when four anchor bolts exist), who is the naysayer to my judgement? If a PE is hired to check the design/calculations, does the rounding and arithmentic become the focus or an independent method used producing similar member sizes and connection reactions? How many load combinations are modeled or can just the critical ones suffice? Are serviceability criteria checked as to vibration, deflection, drift limits? Probably not and definitely not if not included in calculation package.

 

[HgTX]

I'm with frv. I was taught not to go up a whole beam size for a miniscule "overload"--but that I'd better be prepared to defend that decision if I ever find myself on the witness stand. Whether one's comfort margin is 1%, 2%, 5%, that'll vary.

Nothing that went into that calculation to begin with is known to that close a degree. The stresses aren't that well known (other than dead loads, they're not even calculated but standardized), the various load factors are committee compromise based more on "feel" than anything else, the material properties themselves are based on minimum requirements rather than what is likely to exist in the actual material, etc. It's all, to steal a phrase from somewhere else, a wild guess carried out to two decimal places.

Hg

Eng-Tips policies: faq731-376

 

[Teguci]

radair - Read it again.

If the code was suggesting that it was OK to add a 5% load to an existing member that is currently at 95% of its allowable capacity then this section is unnecessary.

Answer the following questions:
1. If an existing member is loaded to only 50% of its allowable is this section saying that you can only add 5%?
2. If this section does not give you the ability, per code, to go to 105% of allowable, what is this section saying?

As for going the final step, you pointed to it in your response. If the original structure was designed to 100% capacity, then this section says that, per the code, you can go to 105%.

I understand that doubling the loads on an existing structure so that the structure is at 105% overstress is not exactly the correct interpretation but a structure (such as a transmission tower) is usually pretty close to 100% utilized before adding todays cell phone receivers.

 

[frv]

HgTX-

That's a great line.. Hadn't heard it before, but it is perfectly applicable.

I had had to contain my laughter when someone on this thread earlier stated that allowing a 5% over stress is "bad engineering". I think we sometimes forget our own job description and can't see the forest for the trees.

If I have a beam with 100PSF live load being over stressed by 5%, I'll sleep soundly at night.

 

[frv]

Great point Teguci.

I was about to point out the same logical inconsistency in radair's post. I was even going to use a similar example.