Here is my answer to R. Mackey's questions. You guys are awfully suspicious. My interest is an open dialog leading to better understanding for all.
www.cool-places.0catch.com/docs/QuestionsFromRMackey.pdf [link added]
I'll try to get to the other questions as time permits.
We're awfully suspicious because, as you note in your paper, your figures are a wild outlier. It's possible for everyone to have underestimated the mass
and Fresh Kills to have grossly exaggerated about the amount of debris they sorted, but it's not likely. The burden of proof upon you is quite high.
From your responses:
Gregory Ulrich said:
[my comments in normal type, Gregory's responses in italics]
Did you take into account:
Steel trusses on each floor? Included in steel.
But are they?
In your
original paper, you never do quite get around to estimating the steel. You say on Page 2 that
"NIST’s value for the mass of steel used in one tower is 100,000 short tons.3 A simplified approximation based on averaging component dimensions provided by NIST demonstrated that this value is reasonable." But then you slice away 10% of this because it's "below grade." How do you know this assumption is accurate? NIST was vague.
Gregory Ulrich said:
Gypsum and insulation packed around the columns? Superimposed dead loads
No, they are not. You are using the "superimposed dead loads" figure for the
floor outside the core, which contains none of the superimposed load borne by the columns. As I'm sure you are aware, the floor outside the core was quite light in construction, basically intended for open-plan offices only, with all of the heavy stuff
inside the core.
This is the most obvious underestimate. You cannot take design considerations for the outer floors and apply them to the whole tower. You've selected the most bare-bones parts of the structure and tossed the rest.
Gregory Ulrich said:
Windows? Ignored but balanced by over-estimating live loads by 140 tons/floor
(see below)
Utilities? Superimposed dead loads
Elevators? Ignored but balanced by over-estimating live loads by 140 tons/floor
(see below)
I don't see your 140 tons per floor accounting anywhere in your original paper, nor do I have any reason to think this is accurate.
As for utilities, rubbish. Where is plumbing? Heating? Backup electrical systems? Water weight alone is going to be at least a thousand tons. None of this is accounted for in the floors outside the core, because they weren't outside the core and weren't usually borne by the floors.
You'll note that on the beam-framed floors (e.g. mechanical floors), you have superimposed loads of 55 and 75 psf. That's seven to ten times what you used for an estimate. Still think you can ignore the core?
Gregory Ulrich said:
Sorry, the reference should be NISTNCSTAR1-1A. The scanned design document is on page. 11. It is important to remember that all of the design (code) loads have safety factors. For example, the design dead load for the concrete slab is 36.5 psf when the actual load based on concrete density is 18.2 psf--nearly exactly half.
This is abuse of safety factors. There is a little bit of conservatism built into the design load, but the safety factor is a multiple of the design load used to size needed capacity, not something that's built in for you. The design load will be fairly representative of actual load assuming the structure is built to spec. This alone may explain why you're off by a factor of 2 compared to other studies... coincidence?
Gregory Ulrich said:
Steel floor framing is dead load not superimposed dead load.
Is it?? Let's take a look at your paper:
Gregory Ulrich said:
Total Mass
The total mass is 279,000 short tons or 254,000 metric tons.
Table 4: Mass above grade
Component .......................... Mass (short tons) .......... Mass (metric tons)
--------------------------------------------------------------------------
Concrete floor inside core area ............... 29 400 .......................... 26 671
Concrete floor outside core area ............. 56 600 .......................... 51 347
Structural steel ................................... 89 416 ........................... 81 117
Live-load inside core .............................. 8 075 ............................ 7 326
Live-load outside core ........................... 38 850 .......................... 35 244
Superimposed dead-load ........................ 17 600 .......................... 15 966
Total mass above grade ....................... 239 941 ......................... 217 671
You may treat that dead load if you wish (I wouldn't, these are relatively temporary structures), but you haven't accounted for it in your table. You don't total dead load, just sum the concrete and structural steel.
What else do you think you missed?
Gregory Ulrich said:
Based on the safety factor of two as validated above, essentially I have overloaded the floors outside the core by 4 psf over an area twice as large as the core. If we take that extra load and put it in the core we are up to an actual load of 16 psf which corresponds to a design load of 32 psf. This is roughly equivalent to your popcorn, beam enclosures, etc. I’ll try to evaluate the sensitivity to this factor in my next version.
Once again, you're misapplying the safety factor. Do you honestly think the superimposed dead load, as-built, only came to four pounds per square foot?
Your bare-bones calculation has no margin to absorb these kinds of errors.
Gregory Ulrich said:
In the meantime, let’s take a look at the loads. What loads are there outside the core?
This is office space: moveable lightweight partitions, desks, chairs, computers, maybe a file cabinet. Think of a cubicle 5’x5’ = 25 sq ft. Using my value 25 psf, we are up to 625 lbs for a cubicle. This could include 100 lb desk, 100lb partition, 100lb file cabinet/books etc, 25 lb computer, 200 lb person, and 100 lbs left over. Then we have corridor space probably a minimum of 10 sq ft per cubicle (another 250 lbs). Are you getting my picture? Now I have 350 lbs extra per cubicle with a possible 800 cubicles per floor. This leaves 140 tons per floor for other stuff (i.e. safes, mainframes, etc.).
OK, so here's the 140 tons. And I do get the picture -- you're low-balling again. Paper alone is around 400 kg per cubic meter, and you're only allowing 100 pounds for file cabinet and bookshelves... the full filing cabinet alone will weigh much more than that.
Gregory Ulrich said:
NIST doesn’t claim any weight! Also you are lumping everything together here. It is only the live-loads for which I make this assumption and this is well motivated above. Don’t forget, my PE is the same as FEMAs which implies the same mass. I’m pretty sure that code is two times the actual expected weight this is why the live-load reduction is allowed.
No. The live-load contribution is only part of the structure -- by your calculations, about 20%. The live loads alone will not dominate your overall conclusion, which is the Towers were built to an actual load double what you calculated.
Gregory Ulrich said:
I never said NIST got the mass wrong. NIST never provides the mass value. They provide very little information regarding their analysis and modelling. If you like to be critical, the NIST reports must be a virtual amusement park for you.
As
Arkan_wolfshade would say,
tu quoque. Anyway, NIST does indeed compute the load -- if you want to be consistent with NIST, look at Chapter 4 of NIST NCSTAR1-2A.
That should be enough. For what it's worth, I have seen another estimate that places a lower mass on the Towers, of roughly 380,000 tons; I will go look for it. It is possible for the others to have estimated high, perhaps anticipating full occupancy -- load of people alone is on the order of 4,000 tons. Nonetheless, I'm quite positive you've underestimated the dead load, the superimposed dead load, and the live load, and furthermore your use of safety factor is incorrect.
