Thursday, August 5, 2010

Reuse and Zombies

I'm not going to be shown up by a shambler.

In my business, you hear a lot of nonsense about old buildings: "They don't build them like they used to." "What a piece of crap." "It doesn't meet code."

Here's the thing about old buildings: they exist independently of our opinions of them. Codes change, but physical reality does not. About fifteen years ago, the U.S. wood code (AKA the National Design Specification from the American Forest & Paper Association) revised downward the allowable stress on nearly every commonly used species and grade of construction lumber. The code had been based on research performed using 2" x 2" pieces, which, it turns out, are not statistically representative of the 2x10s and 2x12s used in construction. A house that I designed before the code change would contain beams that, according to the new code, don't work. Did the house become unsafe because the code changed?

This has happened before. Cast iron columns were common in the 1800s and now are excluded from structural use. Ditto unreinforced structural concrete and concrete columns reinforced only with longitudinal bars.

Old buildings are full of structure that we no longer use or now use differently, but that doesn't mean they're unsafe. It means we have to understand how they actually work rather than blindly applying the latest codes. Most old buildings, if they're in fair or better condition, have greater load capacity than modern buildings because they were designed using more conservative codes, rules of thumb rather than codes, or both.

Engineering analysis is about creating a model of reality that can be easily calculated, because actual reality is too complicated for efficient (read: economical [read: I have to charge fees that clients will pay so I have to stop analyzing at some point]) analysis. But the model is not reality and the code is not a building. Analysis of old buildings for reuse more closely resembles the trial-and-error work of Dr. Gregory House than it does that of the Bernoullis or Euler, even though the Bernoullis and Euler actually created some of the mathematical tools we use.


Rusty Shackleford said...


zombie rotten mcdonald said...

Not to be shown up by an engineer who confuses 5/9 and 9/5:

A while back, I was discussing a potential rehab project with a client. The building had been analyzed as N__B said, using modern methods and codes, and the engineers told the client they had to install steel x bracing throughout the building to maintain lateral stability.

Now, this is a five story building in a low-intensity earthquake zone, and the building was constructed in 1897.

As Ned points out, it was not really practical to analyze the building using contemporary methods; to be safe, the engineers naturally used the lowest estimates of material strength and structural systems, and the most stringent interpretation of the structural requirements, and suddenly a small building needed 1 million dollars in structural work, which would also basically make it unusable.

And yet somehow, the building had stood for 100 years with no appreciable movement or visible distress.

We wound up converting it to condos without the structural work. Still hasn't fallen into the swamp.

N__B said...


Happens all the time. What's particularly annoying for me, as an engineer, is that our design and analysis are based on empirical data, but when you try to use empirical performance as an explanation some jackass will accuse you of not being "a real engineer."

zombie rotten mcdonald said...

some jackass will accuse you of not being "a real engineer."


zombie rotten mcdonald said...

wasn't the downgrading of modern wood standards also due in part to the nature of modern structural wood? So much of it is accelerated growth, and of much lower density than historic woods, or even from 50 years ago.

Sorry about the brain, Rusty.

N__B said...

Some, but not much. The commonly used pine species and grades are not much weaker than before. The big issue is that the old test samples weren't representative: you can't have a 4" long knot in a 2x2.

Smut Clyde said...


zombie rotten mcdonald said...

that was a bit obvious, don't you think Mr. Clyde?