The 9/11 Forum

Major_Tom wrote:Femr, in the video can you see that flash near the south, 95th floor just a second before the 98th floor starts to move?

http://www.youtube.com/watch?v=b_f3d-gETnk&feature=related

Yup. Seems to *flash* twice. It's at the point where the direction of E-W ejecta changes direction from N-S to S-N. It's visible on other videos.

Anything?

Very difficult to say. The video is such poor quality. Will need to correlate to video from NW. It could be fire being blown out. A longer view (even worse quality)
http://www.youtube.com/watch?v=mUaT09OSx_0
Wish someone could get hold of a decent quality West view.

Evidence for bolted connections for all core column H beams in the WTC1 collapse initiation region:


From NIST NCSTAR 1-1C
http://wtc.nist.gov/pubs/NISTNCSTAR1-1C.pdf


From pg 48 of a WTC1 inspection

Core corner columns inspected from office area floors—Core corner columns were inspected at all four corners of the 92nd to 93rd floors and at three corners at the 77th to 78th floors. All seven locations were inspected by removing the five or more layers of gypsum wallboard around the splice locations, which were 3 ft 0 in. above the finished floor level.
On the 92nd to 93rd floors, all four splices were bolted connections. No defects were found at any of the bolted splices. Welded splices were found at all three locations on the 77th to 78th floors, and were tested using the dye penetrant method. Results from the tests showed that there were no surface cracks in the welds. Steel plates, bolts, and welds that were visible were found to have only light surface corrosion.

Pg 39

Core columns inspected from office area floors—Quality Assurance personnel inspected core corner columns at eight splice locations (see Table B6 in the survey report for locations). Seven of the splices were welded, and the welds were tested using the ultrasonic and magnetic particle methods. The eighth Chapter 3 Draft for Public Comment 40 NIST NCSTAR 1-1C, WTC Investigation
location was a bolted splice connection. No defects were found in any of the welded splices. Steel
columns exhibited light surface corrosion, as did the bolted splice on the bolts and splice plates.

The pg 48 quote verifies that all four corner columns on floor 92 were bolted, not welded.


We can also look at the NIST steel inventory for bolted splice plates. From NIST NCSTAR 1-3B:


C-65 (s-8) (904A: 86–89) and C-60. Notice the remains of bolted splice plates at the ends of the H beams. (left click to see the full image)




g) C-155 (904A: 83–86), notice the splice plate and bolts at the far end.


h) HH (605A: 98–101), This splice plate is from the 98th floor of WTC1. It is held to each beam by only 4 bolts in each plate.


c) C-30 (1008B: 104–106)




C-65


C-80 (WTC1 603A, 92-95), notice the bolt holes.

Left click to see the full image. The same column below:


C-42(?)

From the NIST steel inventory and excerpts from an inspection report we know that the following CC connections are bolted splices:

The 4 core corner columns 501, 508, 1001 on the 92nd floor and

1008 on the 92, 104 and 106 floors

904 on the 83, 86 and 89 floors

605 on the 98th floor

603 on the 92 and 95th floors


and we see no evidence of welded H beam connections above the 86th floor. It is reasonable to assume that all H beam connections in the collapse initiation area of WTC1 were bolted splice connections.

From the available literature we discovered that there is a horizontal plane on the 98th floor right through the core where all core columns naturally break in two if the connecting bolts are removed. Every 114+ story core column is already "cut" at the 98th floor. All one need do is remove the top bolt connectors and push the top part over a little, taking advantage of existing "cuts" by manipulating (removable) bolt connections.

The structure had many places to fail. The gravitational forces ultimately are transferred to the columns. These forces are transferred via beam seats (angles) on the facade columns or beam stubs which supported the girders surrounding the perimeter of the core. However the angles and bolts may have survived while parts of the trusses failed, such as the bars or the top chord. Or the trusses may survive while the slabs break apart and drop through the trusses. Ultimately there were thousands of tons of overloading of the structural connections - factors of 10s or 100's of times beyond their yield strength.

My hunch is that bolts failed first, sheared off. If massive loads dropped on the floor and the trusses collapsed they would pull the ends and likely shear the bolts before the deformed or pulled the facade in. To think the bolts held and the columns box sections bent or that the bolts holding the column spandrels or the column to column connections failed BEFORE the bolts of the truss seats is pure fantasy.

When the upper sections began to tilt the same shearing of bolts at the column side in tension took place. These splices were not intended to see the shear loads that was put on them. Welds probably failed too - even full depth welds as these splice plates were stretched apart and cracked from the tensile stress.

Would the facade stand with the floors gone? Perhaps for a bit, but the horizontal shear would torque and twist the bolts in their joints as well and it too would collpase. The avalanche of the floors however provided enough horizontal impulse to break enough of the connections to topple the facade away from the building.

The core column joints were not staggered as the facade was. Every third floor began a new series of core columns with 47 splices.

I'm going to use this thread to group CC connection info.

NCSTAR 1-2A Draft, fig 2-2

That's so cartoon like with the scale completely misleading. First of all many of the columns were I shapes 14" wide flange sections. Is the reinforcing plate on the web or flange or both and on one side or both. And what is the significance of the double dashed vertical line in the center of the column? Is that intended to show the web beyond meaning that the reinforcement is placed on the exterior side of the flange?

Note in the plan from this site http://wtcmodel.wikidot.com/cc703, the column (what does the L mean) above, the one depicted in the NIST cartoon doesn't even HAVE a web, but is two separated plates. (I've never seen a column constructed of 2 separated plates with no web to couple them let alone one that was 3 stories tall which supposedly has lateral reinforcing beams at each floor level.)

Methinks that there is a lot of false information out there.

Oh! I forgot these webless WF columns were used so that could fit some nifty lateral beams right between the flanges. So they just stood those 2 - 36' tall 14" wide x 1" thick plates up tack welded to the column below (splice) until they got the lateral steel in place to stiffen them up - no web required! Boy were they innovative!

Sander, around fl 98 they look like the plate at the end of this column. This is our only 98th fl connection specimen.



Not very impressive. Why would anyone shoot a space beam from the dark side of the moon if you can just pop out a few of those?

4 bolts. Pretty thin plates.

I can't see much, but is looks like perhaps a 1/4" plate is attached to the flange running up to about the E... welded... or not. Hard to read this, but it looks like 2 rivets or bolts connected it to the bottom of the flange through a thin plate which ends just before the HH. But there is no evidence of the reinforcing plates in any of the other steel pictured - none in the columns above. In the other photos they show 3 pairs of holes and what appear to be bolts for the splice - no reinforcing flange plates again. Why does NIST have that cartoon?

Note the columns splices occurred 3' above the floor level so with a 36' long column, floors were connected at 9' from the lower end, 21' and the 33'. You can see this in the photos made during construction.

Note also in the photos of the twisted steel examples of a beam stub which are connected to one side (outside the core floor side) of the perimeter core column's flange. It appears that one section was removed in clean up perhaps right at the lowest of the 3 beam stubs (C-30).

Is the above a 9' long section (1/4 of total length of beam) where the first floor of three floors was connected and is that where the web is severely bent and the flanges looked ripped apart? Looks like a possible read.

From the relationship of web to flange this column is a ~12" flange with a ~16 web and that would make it one of the interior core columns such as 703 for example.

Also note that at the four mechanical floors the thickness of the perimeter core columns is always dramatically increased in the mechanical floors. This obviously had to do with the increased loads requiring stiffer moment connections etc.

I wonder if this column is bent over 180° because the antenna plunged down through the core right above this column?

Just from a layman's perspective:

No cutting apart of the WF columns but kicking and bending of the lower end of a column section above the bolted splice connection by kicker charges, so the column above can miss the section below and fall downward.

98th floor of WTC1 for example:

Not to scale pure sketchy.

Interesting to know how much energy for this method would required.

The splice connections column to column were not terribly critical one the frame was properly braced. There was little to no lateral loads on those connections unless the column developed a moment which would induce it to rotate and slide off the column it rested on (spliced to) or supported (spliced to).

The bracing beams provided strength to the columns by reducing its effective unsupported length. Obviously in the process of erection the columns has to be reasonably erect and stable so that the bracing can be secured with moment connections. Once all the bracing is in place the column is restrained from moving laterally.

One can see how insignificant the column splice are if you look at the columns at the base of the core, where they had wall thickness 5" and could not be bolted or have full penetration welds. But then again tipping a 54 ton column over during erection would take quite a lateral impulse.

Breaking the column to column connection up top would not required powerful charges. MOVING the column laterally... at the connection/splice which is loaded is a different matter. Here is were a diagonal cut would use the load to cause the upper column to displace laterally down the "inclined plane" of the diagonal cut. However the column is braced in at least two direction core corner columns and 4 for those in the center of the core and 3 for those in the perimeter of the core. The OOS floor system would also restrain the column from moving out of the core.

We haven't look much as actual mechanisms to fail the columns. It appears to me from a cursory look at the horse show and other buckled columns that those failures are not the initial column failures but failures which occurred during rapid progression and sever over loading past the yield point of the column... which likely was weakened somewhat from the heat.

What we need to find are column failures which don't appear to be from buckling which resulted from load redistribution. To my eye all the twisted and bent columns are evidence of the progressive column failures but not of whatever caused the failure to progress and the loads to be redistributed.

If the facade was not able to carry its portion of the OOS floor loads the floors would then be cantilevered off the core... but not for long as those truss seats, trusses and slab themselves would not "work" as a cantilever and likely "break off" at the core seats. Of course a local facade damage is a lot different from losing a substantial portion of facade support for a given floor. We didn't see extension loss of the facade around and entire floor so it seem unlikely the facade failure UPLOADED the perimeter of the core. I don't buy the sagging floor trusses did in the facade.

By the way the facade columns were connected to one another by 4 - 1" Ø bolts and their spandrels were also bolted with 2 rows typically of with something like 24 - 5/8" Ø bolts. This was able hold the panels in sheets until the weight sheared the heads off the bolts on the spandrel connections and the column to column splices. Way too much leverage for the bolt heats to resist once the panels began tipping.

This all looks like the designers didn't think about these sorts of possibilities and understandably so. But once events occurred and there began a redistribution of the enormous gravity loads... all bets were off that this structure could perform under THOSE conditions. And it didn't.