Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror

Hurricane Simulator to Destroy Full Size Building 162

Posted by Hemos
from the ka-whump dept.
Anonymous Coward writes "This is a shameless plug, but I thought Slashdot readers might be interested in the hurricane simulator system the company I work for (Cambridge Consultants) helped develop for the University of Western Ontario. The BBC article is light on the kind of technical details Slashdot readers enjoy, so here are some titbits. The servomotors for the 100+ valves are controlled over an IPv4, gigabit Ethernet network connected to an Athlon dual-core AMD64 PC. The entire real-time control system runs on this machine, utilizing well above 90% of each processor core, and roughly 30% of the network capacity. The sampling frequency of the control system places a huge demand on the machine, with about 70,000 context switches taking place every second. Yes, it runs Linux. "
This discussion has been archived. No new comments can be posted.

Hurricane Simulator to Destroy Full Size Building

Comments Filter:
  • by technoextreme (885694) on Monday June 26, 2006 @07:15AM (#15604598)
    I have doubts that you can accurately simulate a hurricane without the space around the house. Some damage is done directly by the wind, yes, but there's a LOT of damage that is done by the wind blowing things into other things and weakening them.

    Yes but scientists have all ready been firing 2X4s directly into different structures in order to test this. Its a lot easier than trying to directly test the effects of wind.
    As for being "perfectly repeatable", I have doubts for that as well. That assumes that you could build the exact same house over and over. The article even states that the placement of the nails in the house matters, and I can't see anyone being that perfect.

    Yeah. Its odd how some scientist can say a measurement can be perfectly repeatable when one of the major tenants of science is that there will always be variance. Perhaps what he meant is that the experiments will be repeatable within housing code because out in the real world the houses will be met with some variance in building quality.
  • by Anonymous Coward on Monday June 26, 2006 @07:36AM (#15604694)
    Hi, I'm the guy who submitted the story, and the engineer who designed the control system and wrote the software. Your CPU stats are hopelessly wrong. The control system is much more than the issue of "blow" commands. It's actually a sophisticated adaptive and predictive controller that requires millions of floating point operations per second, in addition to the overhead of network I/O and context switching. I don't remember the exact breakdown of idle/user/kernel, but I don't think kernel was above 10%. Far from being an abuse of threads, it's actually a conceptually simple and effective way of designing the software.
  • by Froster (985053) on Monday June 26, 2006 @08:57AM (#15605188)
    I figured I'd chime in here as a Western Engineering Student, who had Prof. Kopp last year, let you know what he was up to in the Fall. Kopp only taught the second half of my course because the first half of the year (during Hurricane season) he went to New Orleans to study the devistation.

    This project isn't meant to make a perfectly hurricane resistant house (though, you could try based on the results). As far as I know, the aim is to find what little things can be done to the average house to improve the chances of survival for the house, or at least the people in it. In the example of nailing trusses to the walls of the house, anyone who's actually been there to see or nailed a truss can attest to how weak that connection can be, and one possible change is to mandate exactly how the trusses need to be nailed, and perhaps develop a new nailing plate to ensure that the placement of the nail is exact each time (if there is a steel plate on each truss with only one hole, you know where the nail is going).

    Also, for anyone wondering "why Western Ontario?", UWO is home to a very well respected wintunnel lab, which has tested many very well known buildings (Athens Olympic Stadium, CN Tower, numerous tall buildings in China to name a few). You can take a look here: http://www.blwtl.uwo.ca/Public/Home.aspx [blwtl.uwo.ca]
  • by Anonymous Coward on Monday June 26, 2006 @10:42AM (#15605922)
    You don't need to run simulations yet. There are still plenty of real world data to collect before you can adequately simulate any hurricane. There are tens of thousands of buildings of every type here on the Gulf coast that can be assessed right now. Damage runs the full gamut, from light to catastrophic.

    I live in a FEMA trailer. The western eyewall of Hurricane Katrina passed over my house. Various official guesstimates of wind velocity at the time were high Category 3 to low Category 4 - roughly 130mph or 210 kph. In my neighborhood, houses suffered everything from light roof damage from wind and felled trees, to complete destruction; nothing left but a pile of 2x4's. A few houses were swept away entirely, along with cars, boats and anything else that was not tied down.

    My house is an ordinary 25 year old, rectangular, brick clad, single storey building with a simple hip roof and traditional construction. There are no hurricane straps anywhere in the house. The house structure survived just fine. Not a single window was broken, though most had water infiltrate between the panes. Roof damage was minor, it never sprung a leak, but it has been re-shingled since the storm. What put me into a trailer was the storm surge of 15 feet, over four feet of water in the house for several hours. The interior is still entirely gutted.

    Between damage to the wiring, flooring, drywall, insulation, kitchen and bathroom cabinets, appliances, HVAC unit, furniture and all the rest of the contents of the house and garage, I'd just as soon build a brand new house from scratch. There would be a lot fewer headaches, and not much more expense.

    So what's my point? It's simple. If you'll spend a few month in this area, you'll learn more about hurricane hazards than decades of laboratory simulations. There are too many parameters to get right before simulation results will yield much knowledge.

    IANAE, but from my observation of damage from New Orleans to the Mississippi Gulf Coast there are several points to be made.

    1) Straight line winds are just one hazard. The level of damage varies with wind direction, velocity and duration, amount of rainfall, height and velocity of storm surge, duration of inundation, distance from open water, barriers both natural and man-made, proximity and height of neighboring structures, and tornadoes (there were lots of those embedded in the storm). And maybe other factors too, including luck.

    2) No doubt, building codes play a role in preventing or mitigating wind damage.

    3) Those fancy, intricate gabled roofs that are so popular on all the new McMansions? They suck. I don't care how many metal brackets hold them together. They're mincemeat in a real storm.

    4) You can't do much against flood. High Base Food elevation is the only thing that will prevent flooding. Build high to remain dry.

    5) Most of the stuff you own is located below a line four feet above the floor. The cost of a house's structure is relatively small compared to its contents, equipment, and interior finish.

    6) If you're in the storm's bull's eye, like Pearlington, Waveland and Bay St. Louis MS were, there's not much to be gained by expensive reinforcement of an ordinary house structure. You just can't fully protect against the massive energy that a Katrina-sized storm carries onshore.

    7) Don't trust the U.S. Army Corps of Engineers to protect you against anything.

    8) Don't trust your insurance company.

    9) When a hurricane's bearing down on you, don't worry about the house, get the hell out of town.
  • Re:Yeah right... (Score:3, Informative)

    by drinkypoo (153816) <martin.espinoza@gmail.com> on Monday June 26, 2006 @11:49AM (#15606466) Homepage Journal
    Spelt?

    Yes, spelt.

    spelt [reference.com]
    n.
    A hardy wheat grown mostly in Europe.
    [Middle English, from Old English, from Late Latin spelta, probably of Germanic origin; akin to Middle Dutch spelte, wheat.]

    spelt
    v.
    A past tense and a past participle of spell.

    Source: The American Heritage® Dictionary of the English Language, Fourth Edition Copyright © 2000 by Houghton Mifflin Company. Published by Houghton Mifflin Company. All rights reserved.

    It's both tasty, and correct. (I eat a lot of stuff made with spelt flour, I'm told I have a wheat allergy.)

  • by Buck'sBlog (948076) on Monday June 26, 2006 @04:55PM (#15608991)
    Linux, Sminux. I hate to spoil your fun but the simulator won't work--at least, if you really want to simulate the effect of a hurricane on the structure. The simulator may work for a light tornado, but most of the destructive force of the hurricane would be due to the momentum of the entrained water droplets, not the wind. At about 1/1000th the density of water, air would have about 10 times less force for the same velocity. Good Luck!

The absent ones are always at fault.

Working...