cALCULATION OF YIELD STRENGTH 1

[KAMAL HOSEN]

I’m assessing the performance evaluation of a 10-year-old steel structure in a substation. The design strength for steel was 36000 psi. It has naturally corroded (slightly, seems rust and pitting corrosion, not too much just slightly corrosion). Is there any equation to calculate the current strength of the steel? I don’t want to use 36000 psi, and a steel plate test is not possible at this time.
Thanks in advance.

 

[3DDave]

You are mixing two measures of strength.

36000 psi is a measure of the capacity of the raw material and that typically doesn't change unless it was heat treated and then subjected to red-hot or greater temperature.

The other, which you appear to not know, is the tension, compression, and shear force the structure can resist. This changes when the amount of material changes, such as by rust or wear. There is no formula to tell where and how much material removal has reduced the amount of force the structure can take.

 

[phamENG]

It's impossible to know the true yield strength of the material without testing it. That 36ksi was just the minimum strength of the material coming out of the mill. At 10 years old, there's a pretty good chance the actual strength is much much higher. But no way to know that, as it may also be 36.00001ksi. If you know what the specified material was, then you'll just have to keep using 36ksi in your calculations. (And besides, it's usually a bad idea to base calcs off of a small sample size...the amount of testing you'd have to do in order to get a confident yield strength for design would leave you with significantly less structure than you started with.)

As 3DDave said, you'll have to measure the steel that's there to determine sizes. You'll have to clean the rust from it and look for area of significant pitting. To be conservative, find the worst area and assume the whole structure has lost that much material. If it works, great. If not...find the areas that 'fail' and then go back and figure out the actual section properties at those locations.

 

[RattlinBog]

As others said, stick with Fy=36 ksi. You need to calculate corroded section properties and member strengths.

It's helpful to have ultrasonic thickness (UT) readings or caliper measurements of the remaining steel thickness of flanges and webs. Use a pitting gauge as needed. Then calculate corroded member section properties: Ag, r, I, S, Z, J, etc...depending on the type of member you're analyzing (compression, tension, flexure, etc). Then calculate applicable strength and serviceability values; don't forget to check web and flange compactness in Ch. B of AISC 360-16 (if in the US). You'll also need to calculate demand on the structure, which may be challenging if you have limited reference info. If you don't have access to original design documents, you may need to go through ASCE 7-16 (or whatever is applicable), and apply loads. Finally, check that demands are less than corroded member capacities.

As phamENG said, you can usually take localized corrosion/damage section properties and assume they apply to entire length. It really depends on the member type, too. Is it a beam, column, or tension member? Does it have a lot of shear demand? If it's a tension member, for example, localized corrosion in one spot is all it's going to take to create a weak area. If it's a beam with uniformly distributed load, you'll need to pay greater attention to the flange thickness near midspan and the web thickness near the ends. Don't forget connections.