Thermal Restraint Forces on Tilt Up Panel 1

[lexeng18]

Hi all,

Some background: I am designing a fully composite insulated concrete sandwich panel tilt up wall panel. My situation is a 35'-0" tall panel with restraint at the footing, restraint at 5'-0" above footing into slab on grade, and restraint at the top of the panel to roof steel/diaphragm.

I am trying to back calculate the applied loading values that Thermomass System SC software is giving me with respect to temperature loads. Since I have added a slab tie per my comment above, thermal loads are causing very high moments and restraining forces to my panel.

In a panel without openings, I can come very close to matching the output of the software via hand calculations using the gross moment of inertia to calculate the restraint force at the slab level. However in my specific case I have two dock door openings and I cannot figure out how the thermal forces are modified from the solid panel condition for this case. I suspect that the overall panel moment of inertia is modified (lower) to account for the openings which would in turn reduce temperature loads as compared to a solid panel. However I am unable to back calculate how this is being done.

Would anyone be able to help me reverse engineer these calcs or understand how they may be coming up with their values? If it is helpful, I can post my hand calc to show the solid wall case that matches the software almost perfectly.

 

[r13]

My thinking is simply release the restrain in Y direction of the slab support, if it is not necessarily required to share the weight.

 

[lexeng18]

I have attached my calculation for the solid (no openings) panel that aligns well with the software output.

Anyone have any ideas on how openings that create a varying stiffness panel would affect these calcs, considering moment of inertia is required to solve for restraining force?

EDIT:
had to delete and re post to get attachment to work

 

[r13]

What is your reference for "out of plane" thermal stress? If there is such flow and beam theory is applicable, then basically you have a beam with varying cross section and properties (A, I). You've to take the variations into account in the calculation.

 

[JAE]

I haven't thought through this yet, but it seems as though your 8d in the denominator over-estimates your delta bow amount.
The 8d denominator presumes that you have a purely simple span - 30 ft. long. In reality, you have something other than that, possibly, as the 5 ft. length below the slab is "sort of" fixed as it tries to move against the soil. This paper (

https://digitalcommons.usu.edu/cgi/viewcontent.cgi?article=8241&context=etd)

shows a 32d in the denominator for a fixted-fixed condition, which you surely do not have.

So a number between 8d and 32d is maybe the correct number - but what that is I'm not sure - have to think a bit on it.

Another view is that you should apply the thermal difference to the 35 ft. - get a deflection, then back calculate the load that would be require to push the 5 ft. point back to zero. That would use an 8d in the denominator but it might require an even higher load than what you have derived so far.

 

[JAE]

This link has some brief info on thermal for sandwich panels (

https://www.pci.org/PCI_Docs/Design_Resources/Misc/Sandwich Wall Panels Guide.pdf).

Maybe some references in the back could help you.

 

[JAE]

...also....your Ix is shown as 512. That is for a solid 8" panel, not a sandwich panel.

 

[r13]

Haven't check the entire cal. But it seems the OP is using consistent displacement method to solve the reaction at the slab level, thus 8 should be right.

Response retracted due to lack of understanding on this specific topic.

 

[JAE]

Well, for 30 ft. it is definitely not 8 as I explained above.

 

[lexeng18]

JAE, thank you for the responses. I will give both of your suggestions a try in excel. I can set up the equations and try all values between 8-32 in the denominator to see if they converge to a solution. I can do the same thing with your other suggestion by solving some beam formulas.

To your point about the force being larger by solving for the force needed to move the SOG point back to 0 while considering thermal bow on the full 35'-0" segment, this could be offset by using a reduced moment of inertia. Maybe they are using cracked or effective section properties? Gross makes the most sense to be but I'm not sure.

I will follow up to this post.

 

[JAE]

If there's some substantial axial load on the panel, that can add to the cracking moment value a bit.

 

[r13]

I think you have made mistake in "close the gap" method and over estimate the restraining force P. IMO, you shall estimate the deflection use the length 35' in equation for L, then get the deflection at 5' above the end, this is the gap the force needs to close, not the maximum simple span deflection in the mid-span.

I've not reviewed the linked paper fully, but I think it addresses moment of inertia for the sandwich panel, and how to estimate the moment. Please have a look. Link

 

[Brad805]

When you say "fully composite" do you mean 100% composite action? Do you have the manual for the Thermomass SC software? There is a description of the modeling technique behind the input. It is basically a vierendeel truss analogy where they have tried to match the stiffness parameters of the connectors from push off tests. They do not share a lot of the original test data, but Naito has a useful report that might be of interest Link

Another option is to email Edward Losch to discuss this. He or Kim Seeber could both help you as they have been leading the research for a long time.