How to consider foundation loads generated by steel structures designed with rigid supports

[jochav5280]

Good afternoon,

How are other engineers considering foundation loads generated by exposed structural steel structures subjected to temperature loads? Usually the foundation supports are modeled as rigid pin supports in my office, which can produce tremendous foundation loads due to thermal expansion/contraction of the steel, especially at locations where a braced frame exists.

I've heard some people make arguments that the concrete will also expand/contract with temperature, which will drastically reduce the actual thermal foundation loads, but have yet to see a rational analysis to account for this.

I'd appreciate any help on this topic. Thank you in advance!

Best regards,

jochav5280

 

[jochav5280]

Does anyone have any thoughts here?

 

[SAIL3]

it is very unlikely that one encounters an absolutely rigid support....thermal expansion problems usually arise when there is a "straight-line" condition....mom/braced frames have typically enough inherent flexibility to relieve normal thermal expansion.....

 

[jochav5280]

Thank you SAIL3,

I agree with you, but what do I do with you statement? How would I approach justifying reducing the thermal loads that result from an assumed rigid support?

Best regards,

jochav5280

 

[structSU10]

you can model the foundation support as a spring. May need some geotech input for a reasonable spring constant to use.

 

[jochav5280]

Thank you structSU10,

I'm not sure that's the right approach as what we're after is determining how much the foundation will expand/contract in order to determine how much of the structures expansion/contraction will actually be restrained and therefore induce thermal loads into the foundation. I don't believe the geotech would really impact this.

Best regards,

jochav5280

 

[calvinandhobbes10]

If the braced frame foundations are not tied together, the soil would likely give to accommodate any displacement not relieved by steel frame flexibility.

If the braced frame has a tie beam, grade beam, or wall footing, your question is a good one. Exterior exposure probably triples the expansion dimension. I did a back-of-envelope check for 90 degree F differential and got .25". Without knowing your braced frame, I believe this is on the order of 100-200 kips---which seems within the capacity of a 20" x 20" tie beam. Can you simply size the tie beam for this force?

We had a project charette this week with exterior frames, and that's probably what I'd do. Tie beams also help with drift since they take foundation rotation and let you fix baseplates.

 

[JStephen]

Good question.
One solution:

http://www.pipingtech.com/products/slideplates.htm

One solution: Attempt to determine actual temperature differences.
One solution: Intentionally introduce some flexibility into the steel structure design to reduce the forces.

 

[MotorCity]

Most foundations do not expand and contract since they are below grade and are subject to very small, if any, change in temperature. Depending on the location, the ground temperature remains fairly constant once you get a few feet below grade.

 

[jochav5280]

Thank you calvinandhobbes10,

Agreed. Most of our structures sit on a single mat foundation, so there is not flexibility within the foundation to absorb thermal effects. I've had some very large and long braces that have developed much higher loads than you had even mentioned based on an assumed rigid support. We don't usually design the foundations, so I've just reported the large thermal loads, however, if we did, I'd try to do everything I could to not require support of such large thermal loads. Our industrial equipment is outdoors most of the time and in mountainous regions where thermal variations are quite high.

Thank you JStephen,

Yes, I'm aware of slide-bearings to relieve thermal expansion, but you can use these beneath your lateral force resisting system as it would require too much of the force to be resisted by too few columns.

As for your second solution, that's what I was really after in this post, to see if anyone had a specific method for determined the temperature differences, although, I don't believe they will be much to write home about for the reasons indicated in MotorCity's post.

What would you third solution look like and has it actually been done before? I'm not sure putting flexibility into your lateral force resisting system is a good idea as your braced frame for instance is what needs to be rigid to restrain the rest of the structure.

Thank you MotorCity,

That's been my understanding as well, so I've always just used the resulting foundation loads induced by temperature, assuming that the foundations are rigid.


Best regards,

jochav5280

 

[calvinandhobbes10]

Sketch of your braced frame configuration? Choosing brace configuration that can deflect upwards/outwards may be the best bet

 

[jochav5280]

Hello calvinandhobbes10,

Usually, I'm describing just a multi-story x-braced frame...I hope you can visualize that without a sketch, but let me know if you still need one. With X-braced frames, the braces connect at the base plate of the column, so when the braces expand or contract from thermal variation, this puts outward and inward forces on the foundation pedestals; I curious what brace frame configuration wouldn't introduce this effect; please enlighten me. Thank you!

Best regards,

jochav5280

 

[JStephen]

Most recently, I was looking at a leg-supported vessel with this issue. If the legs are long enough, if they're flexible enough radially and stiff enough tangentially, then all is good.

One problem with some of this stuff is that the temperatures may just be assumed- which is okay- unless you figure out you're paying thousands of dollars extra, then you're motivated to assume differently or actually make some effort to determine the conditions.

 

[jochav5280]

Hello JStephen,

It appears the structure that you described does not have any bracing, which is a situation that I would agree wouldn't generate much thermal shears at the foundation; however, when bracing is added, which is common for our industrial structures, the thermal expansion/contraction of the braces lead to high shears at the foundation when the supports are rigid.

Our temperature variations for industrial projects are almost always defined by our client, so we must use them. Since our structure are normally unenclosed and exposed to the elements, we're forced to design for the expansion from the high temperature and contraction from the low temperature, although not the entire range, (we assume a calibration temperature of 20-degrees Celsius as that is the calibration temperatures of tape measures and the temperature at which steel is supposed to be the correct length called for on design documents.)

Best regards,

jochav5280

 

[calvinandhobbes10]

Here's how I would look at a sample X-braced frame for foundation thermal thrust. Haven't done it before so critique welcome.

This is the first iteration of what would be a stiffness-dependent load distribution between the first-level tie beam, upper brace, and foundation tension. This iteration puts it all in the foundation so I believe it is conservative for foundation design. I'd also design the first-level tie beam for the same axial. I used an HSS10x10x1/2 for brace area, no idea what your dims or sizes are.

If instead of X-bracing you do single-diagonal braces, all the thermal movement differential is taken by bending of the far column (the one with no brace attaching to its base). Foundation can stay as-is, steel frame can expand/contract, and the only internal stress is the column bending (still foundation thrust, but much much less).

You might consider JStephen's approach and use the lowest 6' say of all columns in bending to absorb the thermal movement. Again, no idea what your loads or dimensions are. Concrete pedestals may also be useful since your steel is exposed.