If you need to attach the panels together at the corners to pick up extra dead load, by all means do it.  You are right as far as I am concerned regarding temperature effects.  I wouldn't be worried about the temperature effects as much as the wall failing by overturning.  How much of the intersecting walls do you need to develop to achieve the overturning FS of 1.5?

Mike McCann
MMC Engineering
Motto:  KISS
Motivation:  Don't ask

Each adjacent wall panel, plus a couple extra kip that will be taken care of via the shelf angle and the rest of the wall. To only use the foundations I would have to really oversize them and then come up with an ugly tension detail at the base of the wall for 38kip. Yuck.

I figure one 20ft panel still leaves the rest of the wall free to expand, better than plenty of built structures I have seen from an expansion joint perspective.

a2...if you treat your critical panels as the base or starting point for the thermal and shrinkage effects, I don't think you have a big issue there.

If your critical OT is bi-directional, then your connection orientation is only critical at the inside angle, but even then not a big problem.  If uni-directional OT, then take advantage of nesting the panels at the corners....it will help one side anyway!

I think your approach is fine.

Can you drop your footing depth a bit and get some bending advantage from the higher location of the pour-back strip?  

a2mfk:

You're primarily picking up shear parallel to the length of the weld joint, that is the height of the wall, to overcome overturning, right?  At these corners, inside (right corner) and outside (left corner), these joints are fairly flexible w.r.t. the in plane wall movements you are concerned about aren't they, any movement in the long walls will just open or close the angle btwn. the two panels at each corner, correct?  Maybe pay some attention to which panel passes the other, or which panel edge butts into the face surface of the other panel, so that the primary movement causes the joint from the weld to the outside surface of the panels to open up (i.e. need more caulking on the outside) rather than pinch closed and pry on the weld and the plates.  Maybe actually leave that through wall joint wide, with caulking and backer rod, etc., to allow some rotation in the through wall joint, to allow length change in the long wall sections.

On the inside corner (the right corner), use an angle formed at 93-94° included angle, and at the outside corner (on the left) use a pl. bent around the corner, but formed at 86 or 87° included angle; I'm trying to keep the steel pl. surface nearest the conc. panel surface away from the conc. surface a small amount, so that if there were some angular change at the corner some movement could take place without prying on the welds.  I could try some sketches and see if I can send an attachment if you wish.

a2mfk:

The descriptive phrase I was looking for was, do you want your corners to be able to act as hinge, in plan, for a few degrees of movement, without prying on the welds or plates in the conc. panels.

Thanks guys for all the responses. I have so many other details to work out that I need to move on with this one.

The drag strut force is in both directions, so that is why I have to connect both panels to the ends of the wall. Since each panel is only 20 feet and then there is an expansion joint between it and the next panel, I really don't see why expansion/contraction would be an issue. I don't have a formula in front of me but 20ft of concrete cannot expand all that much in one direction....

a2...you're right.  Exp/Contr not much of an issue and any torsion created by such is probably too small to worry about.

A2mfk,

Really, you will only be able to pick up half a panel at each end with this method.

Another way to engage more weight is to use ribbed grout tubes at the base which have a rated tension capacity and can therefore engage the weight of the footing. If properly designed, the strip footing can also help to engage more than the half panel each end.

csd- yep, I know... I have considered just about everything, but its late in the game and a panel connection is practically a typical detail and the weight is there. The footing contribution was much less significant and a lot of work and detailing to make sure the connections would work.

Thanks to all. Its gonna be a rough 3 days till Monday...... And then I just realized tax day is today, I've been so buried I knew it was coming up but... Ooops. Extension time.

A bit of unexpected good news for you on an apparently not-so-good day at the office. Tax day is April 18 because today (4/15) is a holiday for the IRS in DC ("Emancipation Day").

Nakamura, in Hollywood it is Never tax day..., you know, Bob Fosse's ... "even the orchestra is beautiful!".