Roughly, There are a few things that are generally applicable for earthquake performance.
1) The period fundamental period of a building is roughly .1 sec/floor. (+-, OOM) Earthquakes tend to have an energy peak in the 1 second range. Houses and skyscrapers have an advantage here, 5-10 story apartment buildings, no. (Houses move with the ground, it hits higher modes of skyscrapers, but apartment buildings resonate)
2) Unreinforced masonry or lightly reinforced concrete sucks in an earthquake. Very low tensile strength, which leads to crumbling under shear loads. Combine this with 5-10 story buildings and you have a death trap. Reinforced concrete helps, but it takes a _lot_ of steel, so much that you have difficulty placing the concrete. It's mostly for containment, so that you have this core of concrete wrapped in a steel jacket that can retain enough strength that the building doesn't collapse. It's still a write off after the quake, but hopefully you haven't killed anyone.
3) You want relatively consistent height vs stiffness and stiffness vs orientation profiles. One weak floor (e.g, double height ground floor with columns) concentrates the motion there, and with that the forces. One of the California quakes (Maybe Northridge) showed that the typical for the time arrangement of 3 walls + one open side for parking on the ground level, + 2 stories above was a really bad idea because that one weak wall doomed the buildings. You can't do that in California anymore.
4) Stiffness attracts load. This is a displacement driven regime, rather than a force driven one. If you make something stronger, it often becomes stiffer, and then attracts more load. This can go in a loop. The solution is to make things _really_ flexible, such as using base isolation.
(source, Masters in Civil, working EIT for a couple years in an earthquake region before switching to internet stuff)
3 is called "soft story" and is something San Francisco now requires be fixed or the building is red-tagged - no grandfathering of existing structures is allowed.
Buildings like this are all over Turkey, parts of Central & South America, etc and just as you mentioned the soft story (usually the ground story) tends to collapse causing the building to pancake. It's still popular because it's easy to put parking or a commercial (large open space) on the ground floor for density.
The deadline for remediating multi-family soft stories dwellings in San Francisco was 2020. But this only included buildings with 5+ units, or buildings with a commercial space on the ground floor. There are still many residential buildings, including multi-family buildings (< 5 units), with these 3- or 2-wall soft stories, and there's no mandate in place going forward, AFAIK.
The fix is moment frames, but the typical cost is north of $100,000 for even a single-family home, largely because of the deep footings required, which means tearing up and rebuilding large parts of the foundation and walls. I really wish there were more economical options available. A lot of engineering and regulatory effort went into developing and permitting cheap but highly effective sheer walls, but when it comes to bridging open spaces the only available option (market or otherwise) seems to be a traditional, bog-standard moment frame developed for large, commercial structures, then scaled down.
Contractors don't even offer laminated timber moment frames, only the more expensive steel frames, despite the former technically being available from suppliers and (presumably) capable of passing muster with the building inspector (at least with sufficient cajoling). The premium for steel beams probably only accounts for a fraction of the ultimate cost, but it's an example of the dearth of options, and at $100,000+ those fractions add up.
My 3-story house has a window which compromises the sheer strength of the ground floor back wall, and of course a garage opening in front. Even for the back wall the only options offered by licensed contractors were a deep, concrete footing-anchored moment frame; or removing the window entirely so the sheer wall is continuous, without requiring any footing or foundation work. The distance in complexity between those two options seems huge. In principle there must be cheap techniques for strengthing headers, etc, to provide more than adequate reinforcement, but if available or known they're just not in the average contractor's repertoire, even for earthquake retrofitters, and probably unlikely to pass muster as legitimate reinforcement.
3-5 unit residential buildings are described as the critical "missing middle" in the housing affordability debate in the U.S. I would assume that magic figure holds true in Turkey and most other locales, meaning large fractions if not the majority of residential units are in buildings of that size. The dearth of standard, economical solutions for this middle class of building size in SF makes me think Turkey will find it difficult to reform their system. As evidenced in both SF and Turkey, people will go without if the cost is too high; it couldn't be any other way, economically speaking.
Ouch, it sure seems like reinforcement of a window header, short wall, and the surrounding wall studs would be sufficient... with lots of blocking. You only need enough strength to prevent buckling or deformation right there. I see lots of houses with multiple windows or sliding doors across the back... they seem to treat the vertically continuous wall segments as independent so it's more like several separate back walls. Taken that way you just need enough cross-bracing and reinforced headers/footers no? At least the city is approving those sorts of designs without massive moment frames covering the entire rear wall. (My house fits this with the rear wall being 60% window or sliding door and no steel, but the walls are 2x8 with a lot of blocking and diagonal cross-bracing; the perpendicular internal shear wall connected to it is also built with 2x8s).
Garage doors are as difficult as ground floor parking. They often take up the entire building frontage so you need significant reinforcement.
My house has a steel moment frame set back from the door opening half the depth of the garage because it has to support a 1.5 car width span and the primary sheer wall for the second floor directly above. The house pre-renovation had none of that, just toe-nailed pillars like lots of SF garages which would have made that part of the house vulnerable for sure.
> The deadline for remediating multi-family soft stories dwellings in San Francisco was 2020. But this only included buildings with 5+ units, or buildings with a commercial space on the ground floor. There are still many residential buildings, including multi-family buildings (< 5 units), with these 3- or 2-wall soft stories, and there's no mandate in place going forward, AFAIK.
The R-2/R-3 neighbors of the city are filled with non retrofitted soft-stories indeed!
> The fix is moment frames, but the typical cost is north of $100,000 for even a single-family home, largely because of the deep footings required, which means tearing up and rebuilding large parts of the foundation and walls. I really wish there were more economical options available
They do have solutions (Simpsons makes steel strongwalls that are a lot cheaper than moment frames to avoid you having to have a shearwall on the entire front wall), and you'll find structural engineers that will do plans without moment frames that'll definitely get accepted by the city. You'll also find the other half of structural engineers only willing to do moment frames that'll cost you 100-200k$.
But what happened in Turkey is probably to a large extent not only that the cost was too high, but that the builders skimped on rules, but pocketed the difference.
Now in northern Syria, you could argue the cost would have been too large. The alternative there, in the midst of war, was probably in many cases "build incorrectly, or don't build at all and freeze in a tent".
If I understand you correctly elasticity is the way to go. What sort of building materials and techniques achieve that? (If I want a structure that can still be used after a quake)
eg. Does a floating slab help, can you place the whole thing on big shock absorbers, etc?
> Many seismologists thought the East Anatolian Fault—the one involved in the Turkey-Syria quakes—was likely to produce a maximum magnitude of 7.4 or 7.5. But the February 6 earthquake was a 7.8—about four times bigger on the logarithmic scale of earthquake magnitudes. So it is possible that some structures built to code in Turkey may simply have experienced more force than they were built to withstand, Taciroglu says.
There were probably some such cases, anyway this is as well interesting:
The city that didn't collapse: How Erzin became a haven from Turkey's earthquake - Residents and officials say Erzin suffered no deaths and saw no buildings collapse, and they credit a long-standing policy not to allow construction that violated the country’s codes.
Erzin was the sole district of Hatay Province to have none of its buildings collapse in the aftermath of the 2023 Turkey–Syria earthquake, despite being closer to the epicenter than other cities such as Iskenderun and Antakya which suffered greater damage. The mayor, Ökkeş Elmasoğlu claimed it was due to strict construction of housing; the district mostly consists of single houses, local authorities prohibited unsafe and substandard construction to a much greater degree than nearby areas, and apartment blocks do not have many floors. Even older structures such as houses from 60 years ago survived the earthquake.[4] Additionally, the city had fared relatively well during previous earthquakes. However, engineers and scientists (such as Omer Emre) attributed the town escaping unscathed to geological factors, such as Erzin's relatively higher sea level compared to surrounding towns, and it being built upon harder ground, consisting of bedrock and coarser grains than sand, compared to softer, water-laden sediments like that of cities to the south.
However, I feel we'd be remiss to not investigate the geology of the different places. Are they similar, different, one alluvial one on bedrock? Seeing the iconic olive grove have a 30 meter collapse, I'm not sure what kind of building would survive that upheaval, unless the pilons went deeper than that into bedrock.
I feel we'd be remiss not to read to the end of a paragraph:
However, engineers and scientists (such as Omer Emre) attributed the town escaping unscathed to geological factors, such as Erzin's relatively higher sea level compared to surrounding towns, and it being built upon harder ground, consisting of bedrock and coarser grains than sand, compared to softer, water-laden sediments like that of cities to the south.
This article does a very poor job explaining what its headline claims its going to explain, and even a better article would benefit 1000x from pictures or illustrations.
There are other ways engineers deal with the lateral forces from earthquakes - base isolation (as mentioned in your video), dampeners (as shown in your video), shear walls, braces, and engineered moment frame connections that stiffen the whole structure.
shear walls, braces, and moment frame connections are a lot more common than base isolation and dampeners, especially for smaller buildings.
The regulations and specs are all well and good but they are useless if there is no enforcement. If one can find their way out by bribing officials or by simply building first and getting "peace" or "amnesty" afterwards, regulations will not help much.
This article is skims over the actual engineering details, which most readers might be looking for. These documents from IIT Kanpur in India are an excellent engineering perspective, diagrams included. https://www.nicee.org/EQTips.php
For example "How do Earthquake Affect Reinforced Concrete Buildings? [1] describes the "strength hierarchy" for an RC frame building to remain safe during earthquake shaking, columns (which receive forces from beams) should be stronger than beams, and foundations stronger than columns.
The main problem is the average wage in that part of Turkey is a few thousand dollars a year. In Syria of course is a fraction of that. You just aren't going to get Japanese/Chilean/Californian quality construction.
That also means engineering, management and labor for construction are much reduced in cost, provided you can find domestic engineers working at domestic salaries and management which isn't corrupt.
If labor is all that we have to worry then yes. But building materials prices are almost the same everywhere, so they still wouldn't be able to afford it
Definitely not, a 2x1m plywood board in Tokyo cost ~$50, while the same one back in my hometown of Spain is $10 (despite the average wages being the same in both countries).
I think you might be overestimating Chilean quality/economic circumstances. The big difference between Chile and Turkey is that larger earthquakes than that happen regularly all through the country (every two years at least) - and much larger ones one every decade.
From experience of NZ quakes, wooden buildings seem to survive well.
Aside from the flexibility of wooden vertical load bearing posts, the typical weather board / external wooden slat construction which is widely used in New Zealand allows for the narrow boards to slide independently under the lateral shaking forces. The external walls can deform into trapezoid shapes more effectively - hence a higher threshold for catastrophic damage and better future usability.
Brickwork chimneys, internal plasterwork and the subsurface conditions are another story but a least the dwelling is still functional :/
> the flexibility of wooden vertical load bearing posts
I know you didn't write this, but lots of people think that the flexibility works because it "absorbs" the energy of the earthquake. It does not. What it does is spread a short sharp move into longer but lower amplitude move. The energy imparted is the same.
Ultimately, it's: Ma + cv + kx = 0, the differential equation for damped harmonic motion. M, C, and K are matricies/tensors that can change over time as damage occurs, x, dx/dt, and d2x/dt2 are time derivative vectors of your node DOF (xyz, + 3 rotations). Your boundary conditions are that the nodes connected to the ground have a forcing function xg = f(t).
If everything is linear elastic, you can take the principle components/eigenvectors and get fundamental modes for a first approximation. Once you hit plastic deformation, things go non-linear and get more interesting.
So, Wood:
1) Less stiff, so KX means less force due the ground motion.
2) Less mass, so the MA term is less.
However, wood also tends to be used in smaller structures, so your on the left side of the resonance peak, were first mode structure motion is greater than the forcing function (but not into the resonance peak)
Steel tends to be used for really tall buildings, where the earthquake doesn't excite the fundamental modes, so the ground basically wiggles the bottom of the skyscraper, but the top is nearly unaffected. (the right side of the resonance peak)
Talk to engineers and contractors in Japan and Chile. They have very good recent experience with their buildings withstanding some of the most massive earthquakes in recent history.
1) The period fundamental period of a building is roughly .1 sec/floor. (+-, OOM) Earthquakes tend to have an energy peak in the 1 second range. Houses and skyscrapers have an advantage here, 5-10 story apartment buildings, no. (Houses move with the ground, it hits higher modes of skyscrapers, but apartment buildings resonate)
2) Unreinforced masonry or lightly reinforced concrete sucks in an earthquake. Very low tensile strength, which leads to crumbling under shear loads. Combine this with 5-10 story buildings and you have a death trap. Reinforced concrete helps, but it takes a _lot_ of steel, so much that you have difficulty placing the concrete. It's mostly for containment, so that you have this core of concrete wrapped in a steel jacket that can retain enough strength that the building doesn't collapse. It's still a write off after the quake, but hopefully you haven't killed anyone.
3) You want relatively consistent height vs stiffness and stiffness vs orientation profiles. One weak floor (e.g, double height ground floor with columns) concentrates the motion there, and with that the forces. One of the California quakes (Maybe Northridge) showed that the typical for the time arrangement of 3 walls + one open side for parking on the ground level, + 2 stories above was a really bad idea because that one weak wall doomed the buildings. You can't do that in California anymore.
4) Stiffness attracts load. This is a displacement driven regime, rather than a force driven one. If you make something stronger, it often becomes stiffer, and then attracts more load. This can go in a loop. The solution is to make things _really_ flexible, such as using base isolation.
(source, Masters in Civil, working EIT for a couple years in an earthquake region before switching to internet stuff)