# Structural failure of steel due to heat?

Hi

I was wondering how heat effects the tensile strength of steel. Since there are various alloys under the banner of 'steel', I was hoping for simply a general idea of the effect, if possible.

Here is a quick quote to steer us in a good direction:Most steel has other metals added to tune its properties, like strength, corrosion resistance, or ease of fabrication. Steel is just the element iron that has been processed to control the amount of carbon. Iron, out of the ground, melts at around 1510 degrees C (2750°F). Steel often melts at around 1370 degrees C (2500°F). http://education.jlab.org/qa/meltingpoint_01.html

Given this, is there an equation that we can use to approximate how a steel beam supporting a given weight, would react to increases in heat? I'll take a guess and say that there are a few variables to consider, possibly: Alloy used (of course this is important since impurities and composition determines strength), width and height of the beam, manufacturing process (tempuring), temperature of and nature of the heat source, and finally the weight atop of the beam.

If someone here is expert enough to fill in these variables and design an illustrative scenario, that would be great. My goal is to understand how heat effects the beam's ability to hold the weight and at what temperatures the beam will fail at.

Also, if any civil engineers here know building codes and in particular what the minimum codes are for a steel beam supporting a weight, that would be nice to know. In other words, 'a steal beam must be able to support at least 3x its weight load to be used in a commerical building.' I'm looking for an answer like this.

Thanks.

Last edited:

## Answers and Replies

Q_Goest
Homework Helper
Gold Member
Hi Chaos, I'm not a civil or structural engineer, but I dabble just a tiny bit from time to time. What little I know about it is there are http://en.wikipedia.org/wiki/Building_code" [Broken]that govern their design.

Regarding the affect of temperature on steel, the strength of materials comes from emperical data, not analytical. One of the most common building steels is ASTM A 36 carbon steel. In large buildings the code requires it be protected from heat. http://www.pwri.go.jp/eng/ujnr/joint/35/paper/71sakumo.pdf" [Broken]can help explain why:
http://www.pwri.go.jp/eng/ujnr/joint/35/paper/71sakumo.pdf

Scroll down to page 5 and check out figure 4. This is a graph of strenth as a function of temperature. Note that it only goes up to about 700 C where it stops at 20% of the ambient temperature strength. There will be a fairly sharp drop off in strength somewhere above this temperature as it begins to change phase. Figure 8 a little further on is also very telling.

Last edited by a moderator:
Astronuc
Staff Emeritus
. . . building codes and in particular what the minimum codes are for a steel beam supporting a weight, that would be nice to know. In other words, 'a steal beam must be able to support at least 3x its weight load to be used in a commerical building.' I'm looking for an answer like this.

You might try - Steel Construction Manual, 13th Edition, American Institute of Steel Construction
With this revision, the previously separate Allowable Stress Design and Load and Resistance Factor Design methods have been combined.

Refs on A36 -

http://www.civil.umaine.edu/cie111/images/steel.gif [Broken] - I would expect for room temperature.

http://www.civil.umaine.edu/cie111/tension/ [Broken]

For structural analysis, one needs a temperature dependent constitutive model for the appropriate range of temperature, which would be employed in a finite element model.

As Q_Goest, there are requirements for fire protection of the structural members.

Steel looses strength as temperature increases. Another consideration is load redistribution as the hotter part of a structure looses strength.

Take a look here - http://www.aisc.org/Template.cfm?section=ePubs [Broken]

Last edited by a moderator:
PerennialII
Gold Member
........ continuing the thought about material models, the introduction of time and rate dependent plasticity is taken into account using viscoplastic material models (often referred to just as material models for creep). There are a number of classics when talking about slow rates of deformation, such as the 'Norton' (power law) model with various modifications such as the time and strain hardening variants, quite often see hyperbolic variants used and implemented in FE codes (instead of the power law dependency) and so on (there are a really large number of models available for different materials, micromechanisms they exhibit under differing conditions, if taking into account how material damage during the process affects can multiply the number of models by an "order of magnitude"). For example if you'd be doing an analysis following a 'power law' model, your time-independent analysis would be added with a time-dependent term of form : $\dot{\epsilon} = A \sigma_{e}^{n}$ i.e. strain rate is given as a power-law of the equivalent stress state (the constants being temperature dependent).

Wow. Thanks fellas, its rare to see 3 perfect replies with no filler. Many thanks for such insightful and nicely referenced facts. I shall try to digest it all and follow-up if I have more questions.

thx again!