Tensile Strength Test

[SteelPE]

I am currently working on a renovation project. Building was built in the early 70's with a renovation being done in the early 80's. The owner would like to add some load to an existing steel beam and we were unsure of the strength of the material. The owner decided to cut a piece of the beam and send it to the lab for testing. The lab came back with the following:

Ultimate Strength = 65,500 psi
.2% Yield Strength = 48,300 psi

So, I have been taught to use the .2% yield number which would be 48.3 ksi however, this number is a little odd as it is short of the 50ksi one would expect. Is this normal for these types of tests? I'm inclined to use a value of 48ksi but I am not sure.

 

[jdgengineer]

Do you have drawings that state it is 50 ksi steel? Wouldn't you expect it to be A36 given the era?

 

[SteelPE]

We do not have the drawings. Due to the numerous renovations done, we are not quite sure when this beam was added. The beam works at 50ksi and does not work at 36ksi so just assuming 36ksi is not going to satisfy the owner.

It is slightly more difficult than the explanation, just wondering what to use based upon the actual test results.

 

[EdStainless]

While it wouldn't surprise me to find a location in 50ksi min structural that measured 48ksi, you are now left with a different question.
Since you have no product information, and you do have a 48kis test in hand what assumptions can you make? Are you safe using anything over 45ksi?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[structSU10]

AISC has an article that touches on this issue :

https://www.aisc.org/globalassets/modern-steel/steelwise/30762_steelwise_reno.pdf

the statement that coupons from different areas of the same beam may yield different results is disconcerting. It is a judgement call, and if you are comfortable with one tensile test, then go for it. You could also do a load test program.

 

[MotorCity]

As with any testing, the allowable values are not based on the largest nor the smallest test results but somewhere in between. I would rest well using 48 ksi since it is based on "real" information and not stick my neck out. After all, that was the whole point of the testing, right? Maybe it is really 50 ksi nominal material, and your particle beam fell a little below the curve.....even more reason to use the lower number. In practical terms, will it 48 ksi vs 50 ksi make a difference in your design/analysis?