Appropriate use of Fy 1

[Lion06]

If a structure is designed with A36 and someone made a very BIG "woops" (not me), and the steel is tested and comes back with Fy > 50 ksi I believe it is ok to use A572 or A992 for a second look (provided the other requirements of the standards are met). Is it appropriate to use Fy = say 61 ksi if that's what the test shows or are you limited to a max of 50?

Any opinions?

 

[ajh1]

Reevaluation and initial design can follow two different rules possibly. Initial design is limited to the yield of the grade specified, regardless of what any test reports might show. The challenge with using something other than speced yield in any situation is that while the piece tested may in fact show Fy=61, how do you guarantee that every other piece on the job has an equivalent yield.
That said, most A36 produced today will in fact have a yield reasonably consistent with A572 Grade 50. You can't call them the same thing because one is a carbon steel and the other is a high strength low alloy steel.
I would limit yourself to no more than Fy-50.

 

[DWHA]

What was the mistake? Was the steel tagged as 36ksi yield strength from the mill, and tested to be 61ksi. Or was it tagged as 50ksi steel and tested to be 61ksi? I don't think that I would do anything more than justify a localized over stressed area utilizing the 61ksi. I would not account for it thru out the structure.

Remember, that the actual steel stress form a bell curve, so if there is steel that tested at 61ksi, where is the steel that is on the lower side of the bell curve...

 

[miecz]

I would treat this as an existing building. ASCE's Guideline for Stuctural Condition Assessment of Existing Buildings has provisions for testing materials of existing buildings. I can't find that ASCE specifically says what to do with the tests, but it seems to imply that one should use the test results in a structural evaluation, with no reductions from the tested values.

 

[rb1957]

is there a problem that the material is almost 50% stronger than the minumum spec ?

I guess your code makes assumptions of ftu based on fty ... could you show that ftu of the material exceeds ftu assumed in the code?

 

[slickdeals]

I recently heard at an AISC seminar that recently most of the A36 angles are almost 50 ksi because they are being made from recycled 50 ksi material.

 

[Lion06]

The problem is an existing structure that I am checking was designed with A36, but is SIGNIFICANTLY overstressed (on the order of 200%, even using the newer spec). The material was tested and a majority of the coupons came back with a yield stress around 60-65 ksi and ftu around 90 ksi. Armed with that information, I'm not sure what values are appropriate to use when checking the capacity as is.

 

[rb1957]

so you're questioning the material the structure was built from ? would hardness testing help you out ?

 

[Lion06]

I'm asking if it is ok to use 65 ksi in a design check since that's what the test came back as. I don't know if there are other requirements (such as Fy/ftu ratios, chemical compositions, elongation %'s, etc...).

 

[bridgebuster]

EIT:

If I'm reading correctly, it sounds like you're checking multiple members. If that's the case I would use the actual Fy for the member that was tested and for the others use Fy somewhere between 50 and 60. You could probably justify 60 if there's a reasonable certainty that the steel came from the same mill. I know it makes for tedious work but you'll sleep easier at night. This is not uncommon in bridge work; sometimes we're faced with rehabbing an old structure with no available plans.

If you're not dealing with any fracture critical members elongation shouldn't be a problem. If you need to retrofit members by welding chemical composition should be checked.

 

[JAE]

Well, technically you would gather all your material tests, establish a mean, a coeff. of variation, etc. and then set as fy the 90% confidence value of the scattered plots of the tests....a sort of statistical method I guess.

There might also be an IBC section (chapter 17) or ASTM out there that gives direction to this.

 

[rb1957]

bridgebuster,
i think 61ksi is what the OP needs to pass the structure, and 50ksi is what the tests show.

JAE's right about determining a minimum strength from your population of tests; and your tets should cover each element, preferrably multiple times.

if i'm right about the 61ksi, i don't see how you can increase a test strength of 50ksi ... maybe plastic bending (form factors) can help. but your right that you probably need to check other things, like ftu/fty assumed by the code.

 

[PEinc]

Even if you can show that the higher yield strength helps compensate for the 200% overstress, what about deflections? With much higher load, you should have much higher deflections.

 

[Lion06]

The actual tests have come in around 60 ksi. They will work using that (They were designed using A36). Some deflections are a problem.

 

[miecz]

I still see no reason to use anything less than the (properly) tested yield strength. Be sure to check the connections.

 

[DRC1]

NO Don't use the coupn strength. The yield strength of a beam varies throughout the cross section with the highest yield strngth at the ceter of the web (where it is needed least) to a minmum at the flange tips (where it is needed most) So, at least for mill tests, where is the coupon taken? At the center of the web. Further, A-36 is not common any more for steel shapes. 20 years ago it was, but grade 50 was also popular. Depending on what you could get out of a rolling or in a yard's stock, 50 ksi was often substituted for grade 36. This was considered an improvement and often not documented. Unless you have mill test on the beams, you can not tell if it is A-36 or A572. It is possible they are intermixed. How old is the building?

 

[Lion06]

The building is from the 90's (which surprised me that A36 was spec'd). I don't know where the coupons were taken.

miecz-
Where in ASCE 11 does it talk about what strengths to use? I could only find what tests measure specific properties and couldn't find anything about if you are able to use the full test value from the coupon.

 

[miecz]

StructuralEIT

As I said earlier, ASCE doesn't explicitly state what to use. However, I believe it implies the use of the tested properties.

Article 2.3.3.3.

Both non destructive and destructive procedures may be necessary in order to establish more realistic properties for the detailed assessment.

Also, as mentioned by bridgebuster, this type of analysis is commonly done in bridge work. Article 6.6.2.1 of the Manual for Condition Evaluation of Bridges states

When non-specification metals are encountered, coupon testing may be used to determine a nominal yield point.

I think that if ASCE or AASHTO expected the engineer to use something other than the tested properties, they would have clearly said so.

 

[swearingen]

Don't forget that if this is a high-seismic area, you are required to have a good bit of distance on the stress-strain curve between the Fy you design for and Ftu. This is essential and is assumed to be present by the code's equations. Artificially increasing the Fy used in your calcs (whether it is measured or not) can cause problems in the reserve inelastic capacity of a structure in a major seismic event when using the code. The idea is to prevent collapse...


If you "heard" it on the internet, it's guilty until proven innocent. - DCS

http://www.eng-tips.com/supportus.cfm

 

[HgTX]

I agree with the first answer you got. What's sold as A36 has been rolled to a dual A36/A572 standard since as far back as the time your building went up, so 50 ksi is probably a safe assumption given test results of 60 ksi. However, I don't know that I'd rely on the test result of 60 ksi applying everywhere.

That said, I see MTRs used on occasion to "establish" actual strength of existing structures, and that's treating a whole plate as if it matches the test coupon performed from somewhere in the same heat at the mill. However, the location of the mill test coupon, I believe, is chosen so as to give a conservative result, whereas a coupon from a piece in service will not have been taken from a similar location in the as-rolled plate. Appendix X2 of ASTM A6 has some interesting information about variability of tensile test results.

Hg

Eng-Tips policies: faq731-376