freezingbum

Why do they? I've noticed that they use more structural concrete for their mid rise buildings. Isn't concrete more brittle, so wouldn't they want the structure to yield?

aibarr

*looks around, realizes I should probably answer this*

There's one thing that makes concrete structures feasible: rebar.

The addition of steel with concrete actually makes concrete surprisingly ductile. It's really bizarre... The stuff can flex a lot more without breaking than you'd think. Go YouTube "Tacoma Narrows" and watch the road deck go into cyclic torsion. (Yeah, it eventually collapses, but that's a concrete road deck that's flexing until it *does* go. It's pretty sweet.)

There are a few different failure states that concrete structures go through before ultimate failure, and they're easier to visualize if you imagine concrete structures as being made of something like pretzels (concrete) with an inner core of rubber band (basically the rebar). The first state is pre-cracking action, where it's just the concrete acting. In our pretzel-rubberband model, the pretzel material itself takes all of the force and the rubber band is just chilling, not doing anything. The second state is post-cracking action, where the concrete has cracked but is sticking together and the steel rebar cage is brought into play. (The pretzel has cracked but is providing some rigidity for the rubber bands, which are stretching and giving in response to the forces.) This is a really ductile, flexible state, and it really doesn't take a lot of cracking (just really small cracking, some maybe 1/16th or 1/8th inch cracks and a lot of hairline fissures... which *all* concrete buildings have, to a certain extent) to get to this state. After it's reached this state, it's still more or less okay, and doesn't require major repair.

The third state is where the steel starts to yield and undergo plastic deformation, and that's where these concrete structures start to fail. It takes a lot of energy to get to this point, and concrete structures in seismic regions are designed to go through a *lot* before they get to this state. (In our model, the rubber bands are stretched to capacity, they snap, and the whole thing comes down, but like steel, it'd take a heck of a lot of force to snap those rubber bands in relation to the pretzel that surrounds it.)

So basically, we as structural engineers *can* do it. We've figured out what needs to happen in order to make structural concrete work in seismic areas.

Couple other points... You might be like, well, yeah, it *can* work, but you end up with a ton of cracks after a really big earthquake. I'd in turn answer that philosophically, engineers don't really care about cracks after earthquakes. We try to minimize damage during the smaller earthquakes, sure, but during large earthquakes, we don't give a hoot about cracking. Our goal is *life safety*. So long as the thing stands up long enough for all the people to get out safely, we don't mind scrapping the building. (In order for it to get to that "we don't mind scrapping the building" stage, it'd have to be an 8.5 or 9.0 earthquake... we're talking really big ones, here.)

Why not just use steel instead? Several reasons...

First, concrete's cheap. It's basically a little tiny bit of steel being held together by superprimative rock-glue. The concrete doesn't do a lot structurally except provide some bulk to build out of, and we primarily rely on the steel to do all the cool force-dissipating jazz. Steel alone is far more expensive.

Secondly, the "lead time" on steel is a lot longer. If I have a newly-approved design for a concrete structure, I can buy some wood and get my workers together and start nailing together formwork tomorrow, and I can be pouring by the middle of next week. With steel, I have to send the plans to the fabricator, who has to order the steel to spec from the steel mill, and do all the detailing, and send it back to the engineer to get the details approved, and blah blah blah... Takes months to get that project started. (Once it starts to go, it goes faster than concrete, but for a mid-rise building, concrete tends to win the ultimate race.)

Finally, steel doesn't work all *that* much better than concrete does in seismic situations. There was a *huge* controversy in the welding industry right after the Northridge Earthquake in... I think it was 1994. It was, before Katrina, the most costly natural disaster in recent American history... Hundreds upon hundreds of structural corner welds of beams to columns just flat-out popped. Ruined the columns that the welds were attached to, in a lot of cases. Didn't cause any fatalities or collapses, but the damage to the structures was just tremendous... And it had to do with the fact that the welders used things called "backing bars" to give them a substrate to weld stuff too... Long story short (too late), it was one little tiny detail that was previously held to be a perfectly acceptable thing that ended up, because of the dynamics of that particular earthquake, being incredibly damaging to a wide range of structures. (If you'd like to see some pictures, google image search "northridge weld fracture" and look at some of the things that pop up...)

Anyhow, that was probably more answer than you wanted. Let me know if something needs more explanation!

freezingbum

thanks a lot, that really helped...but...i have a few more questions if you dont mind.

Are there connection issues with concrete frames? I mean it isn't always continuous, is it?

If cost and time and imperfections wasn't an issue then would concrete have any advantage in holding up and safety for mild earthquakes?

aibarr

No problem; glad to help. =)

Concrete actually *is* typically continuous because it's easier to build that way. It's usually poured in different stages, or lifts, but care in design and construction is taken so that the entire thing acts more or less monolithically (in other words, "as one piece", for those of you who haven't seen 2001). The only cases in which we provide separations in concrete structures are in controlled situations to allow for long-term effects like creep (non-immediate, long-term deformation of concrete under constant loading) and shrinking of the slab (concrete has a tendency to shrink for a long time after it's started to cure). We usually do that at things called "control joints," which are usually a good distance away from the columns and critical areas, where the forces aren't so complicated. But yeah, connections typically aren't that big of a deal in concrete structures, like they are in steel structures.

The only sort of corrolary to that is that when you design your rebar layout at all the column and beam intersections in seismic situations, you have to be pretty careful to do certain things like wrap the rebar snugly against one another so it doesn't jockey loose in an earthquake when concrete starts spalling off (you want your steel rebar cage to stay more or less intact, even when the concrete starts to break), and to keep as many of the rebar in your beams continuous through where they run into columns so that if you lose a column, you won't start a progressive collapse... But that's the only sort of area where you have connection issues, per se, in concrete. The connections are within within the concrete, though, and it's not like you have to bolt or weld one concrete beam to another, like you would have to in steel structures. But the concrete rebar connection design issue is a really huge topic of discussion among structural engineers, and it typically goes by the name "detailing." We're always trying to look for better practices of placing rebar within concrete frames so that it works better in weird (earthquake etc.) situations.

On to the second part of your question...

If cost and time and imperfections weren't an issue, would concrete have an advantage over steel in minor earthquake situations? Well, it might, in a lot of cases... It really depends on the earthquake, the structure, what the structure is used for, how it's laid out, stuff like that. Whether to design something out of steel or concrete more has to do with a lot of non-earthquakey factors like vibration control (if you have a hospital with sensitive equipment, steel beams make floors a lot bouncier, which would be bad, say, if someone's OPERATING on my BRAIN with LASERS omg...) or like if you have long, sweeping curves to parts of your building (easy to bend steel gracefully to make architecturally-tasty things like arches or curved-layout walkways, not so easy to build formwork to pour concrete to do the same thing). So, that's what typically controls the selection process.

Bottom line, we can pretty much make anything work okay in seismic situations (except for certain types of unreinforced masonry... bricks or concrete block without rebar... but that's an entirely different discussion) without too much hassle, so long as we know what we're doing (and in California, you have to go through seven years of on-the-job and grad school experience and 34 hours of testing in three or four rounds in order to get your SE... Structural Engineer... license, so the folks who design these things *do* know what they're doing) and we take a little extra time to do things properly.

Good questions! Let me know if you (or anybody else) is interested in hearing more about any of this, and anybody who's reading who also does this stuff and perhaps is lurking, feel free to add anything you think I've forgotten!

Luminaire

albarr: it's like you read my mind. That's almost exactly what I was going to post! ya snooze, ya lose......

ken285

Another major criteria in choosing between steel and concrete is the size of the open spans that you want. Long spans are cheaper when built with steel rather than concrete, so steel is preferred in office buildings, where you want lots of open space. In residential construction, this is much less of an issue, so concrete is the choice material.

I believe this was mentioned earlier, but time is such a major factor as well. You can pour a floor every 4 days (or 2 days in NYC), but steel takes forever to go from the drawing board to the site. Most owners want their building ASAP, especially residential developers, so they can lease/sell the space and start collecting money & interest. In fact, some are so anxious that people start living in the place before the building has been completed. Where I'm working, we had residents move in this past September, but we won't be done for another few months. Time is money, and money is what the owner/client cares about the most usually, as long as it can be built.