Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror
×
Science Technology

Tracking the Cracks 140

Posted by ScuttleMonkey from the broken-news dept.
Roland Piquepaille writes "Israeli physicists from the Weizmann Institute have used a new approach to study how materials break. In a short news release, brilliantly titled "Breaking news", they explain their new method for analyzing the progression of a forming crack. The news release even says that it could have help engineers predict 'exactly how much pressure the levees protecting New Orleans could withstand before giving way.' This method could be used by engineers and material scientists in a vast variety of applications."
This discussion has been archived. No new comments can be posted.

Tracking the Cracks More

Tracking the Cracks

Comments Filter:

  • The trick is... (Score:4, Insightful)

    by Anonymous Coward on Sunday February 05, 2006 @03:27PM (#14646894)
    never design something so it will hold exactly what it needs to stand up against. Unless you're building for suicidal adventurers, people will appreciate headroom. Especially people behind levees...
    • Contrarily, It's fine to design something so it will hold exactly what it needs to stand up against, as long as your aware that what it needs to stand up against is atleast double what you would ever expect it to stand up against.
      • double? I think an order of magnitude is a better assumption - i.e. add a zero

        • Re:The trick is... (Score:3, Informative)

          by AaronPSU777 ( 938553 )
          You may not want to fly anymore. Airplanes are typically designed with a factor of safety of less than 1.5. An FOS of 10 is usually overkill, do I really need to design a bridge so that it will hold ten times the highest load it will ever carry in it's lifetime? In a word, no.
          • Well an FOS of 10 wasn't uncommon in the days of steam, when IIRC correctly the con-rods between driven wheels would be designed to have an FOS of 30+. More a comment on the (lack of) insight engineers at the time had into what makes con-rods break, rather than anything else.

            Returning to 'planes, yes, an FOS < 1.5 sounds about right. Back when I had anything to do with plane design (military), the airframes were designed to support "120% fully-factored load". Whoa, we all gasp, only 20% over the maximu
            • A bit of an aside; For wing attachments on an agile fighter, fer instance, the max load may be based on 9g manouvers which may not even be aerodynamically possible or may never be allowed by the control system. I don't fly agile fighters, I probably never will. I doubt I'll ever be in a situation that warrants it but just in case I never want a control system that disallows actions which might break the machine. I don't care if doing it puts me at greater risk and sure, give me all the warnings you want bu
              • The second Vans RV-9 (as I recall) broke up in flight because a low-time pilot yanked on the stick and exceeded the *insane* specifications on the astoundingly solid design, 12 g's. The wing spar broke. He killed himself and his passenger because he was an idiot. Likewise, it's likely the Airbus that fell apart over New York City four years ago did so because of large rudder movements. Personally, I think you shouldn't be able to snap an airplane in half through overcontrol: I think it should be built t
          • Obviously this plane could have used a higher FoS number... http://video.google.com/videoplay?docid=-640276788 2357542179&q=md80 [google.com]
        • double? I think an order of magnitude is a better assumption - i.e. add a zero
          But I thought adding a zero would just double...
        • Obviously it depends on what's being engineered, but double would be considered overkill in most chemical plants. (I am a chemical engineer.) An order of magnitude would be used in only the most extreme cases. Think nuclear weapons or Yucca Mountain.
        • So if you build a chair that's supposed to be safe for a 400-pound person, you should design it so it will stand up to being run over by a 2-ton truck? Yeah, that makes a lot of sense.

        • I believe the factor of safety for stell structures is in the order of 1.5. As for earthworks the factor of safety can be up to 3. A factor of safety of 10 is not needed. But you DO have to maintain the structures so they are still at their designed capacity.

          I don't have the details about the New Orleans levees, but I honestly doubt cracking had anything to do with it. Such huge works are rarely made of expensive materials such as concrete. It's just too huge. Usually only the sections of levees pro

    • Re:The trick is... (Score:3, Insightful)

      by m0nstr42 ( 914269 )
      never design something so it will hold exactly what it needs to stand up against. Unless you're building for suicidal adventurers, people will appreciate headroom. Especially people behind levees...

      Nobody really does this. That's a standard part of engineering education. Find the exact parameters that you need to work within and then work squarely (and safely) between or above them. For example, maybe the levee needs to be X thickness to withstand a reasonably large hurricane, but at Y thickness the c

    • Re:The trick is... (Score:3, Informative)

      by jeepeagle ( 682756 )
      Henry Petroski's classic To Engineer Is Human : The Role of Failure in Successful Design shows its age a bit, but it's a great read on structural engineering, factors of safety, and failure to learn from the mistakes of the past. http://www.amazon.com/gp/product/0679734163/sr=1-1 /qid=1139177043/ref=pd_bbs_1/002-3727742-0917603?_ encoding=UTF8 [amazon.com]
      • I was going to recommend the same book, given that it's *precisely* about this, but you got to it first. Great book, great recommendation.
      • I'll put in a "me, too" recommendation for this book. Designing with a factor of safety in mind is obvious, but all too often the end result is compromised becasue either the engineers didn't understand the loads imposed or else the builders didn't follow the engineers' design.

        Another great book is "Engineer to Win" by the late Carroll Smith. Even if you're not a racing car nut, this book presents a thorough overview of applied materials science and failure analysis, in some very plain language. Get it at

    • You got to be kidding me...

      FYI, there isn't an civil/mechanical engineer in the world who designs structures to hold the exact value of predicted load. Every structure project is designed with security coeficients which are applied to the the maximum possible load which will be applied in the structure (in certain countries the security coeficient ranges from 1.25 of housing structures to 10 of levees and damns). In civil engineering applications, those coeficients are specified by the country's regulations
      • You apply safety factors to the maximum probable loads, not the maximum possible loads. The maximum possible load is infinite, which wouldn't require much factoring to ensure a safe design.
    • As I understand it, the levees were overtopped by water blown into the lake and reversing up the canal. So the strength of the levees was not the problem. The levees were not high enough for that storm. Therefore, exactly how they broke doesn't matter, once water was washing over them, all bets were off.

  • Plumbers (Score:5, Funny)

    by Mozk ( 844858 ) on Sunday February 05, 2006 @03:29PM (#14646902)
    Plumbers should especially pay attention.

    • Re:Plumbers (Score:2, Insightful)

      by Anonymous Coward
      The people who will probably pay the most attention are those who are developing pressure vessels (especially nuclear pressure vessels). Crack propagation and arrestation are very important to understand especially in brittle fracture or stress corrosion cracking contexts. When you are talking about releasing high pressure steam or radioactive material if a system breaks you tend to care a lot about understanding where it is safe to operate.

      An easy way to break a pressure vessel is to add just a little ch
      • Thats not important, whats important is figuring out who's fault it is ;)
      • Materials 101 (experts please skip post):

        Nuclear pressure vessels are a demanding art form (which is the black art needed until you can fully nail down the science). Radiation embrittlement is caused by (mainly) energetic neutrons displacing metal atoms from their lattice positions. In this sense, it mimics the effects of stress in metals. Since most commercial metal vessels are not single crystals (although they have directionality from the production techniques applied to them) predicting the exact p
    • The people designing pants should also be wary of plumbers. The force of their cracks cannot be denied!

  • Practically applicable? (Score:5, Insightful)

    by onlysolution ( 941392 ) on Sunday February 05, 2006 @03:31PM (#14646907)
    In the article they say taht they have applied their method to a variety of materials, namely plastic, glass and metal. There is a common thread there though, in that all three are higly regular materials. In an earthen levee, or even a contcrete one, the materials used to make are way more irregular than what they have tested their methods with. It sounds like the connection to New Orleans levees is really premature to me.
    • It sounds like the connection to New Orleans levees is really premature to me.

      The other issue is that the New Orleans floodwalls are thought to have failed because the soil beneath them became waterlogged and gave way. Is the model going to work in that kind of a situation?

      • Nope. Cracks develop in materials subjected to tensile stresses or bending moments (where on the outside of the curvature there are tensile stresses). Soils are only designed for shear strength as obviously most soils have no tensile strength (unless they are cemented). Cracks were not the problem in the levees. For the record, most dams are earth dams, too.,

        The problem of piping and erosion is not stress-related. It's all about the pore water pressure and the permeability of the soil (under the leve

    • Yeah. Brittle fracture and soil mechanics are completely different problems.

    • Re:Practically applicable? (Score:4, Insightful)

      by phayes ( 202222 ) on Sunday February 05, 2006 @04:34PM (#14647107) Homepage
      it could have help engineers predict 'exactly how much pressure the levees protecting New Orleans could withstand before giving way.'

      Translation:

      We'd really like in on some of the millions of dollars the Government is spending on New Orleans...

      • Agreed.

        Notice the thought: "It could have helped" ? That implies, given what we now know, we have derived a model that, incredibly, fits the data that has so far been collected. In fact, r=1 for this model which demonstrates how the levees failed.

        Uh huh. And I can predict yesterdays stock price too.
        • I can predict yesterdays stock price too

          I work in field where: A+B+C+D+?=Z

          Z is the observable. We have little pieces of A and perhaps C but the other factors are simply unknown to us. Given that I know Z can I predict Z?

          A large solar flare happens(A), a bunch of stuff happens(B+C+etc...), energetic particles hit the Earth(Z).

          Knowing yesterdays stock price and predicting yesterdays stock price are totally different (just a semantics argument I know/predict?)
      • Millions? With an "M"?

        You know you've missed the mark when even Dr. Evil is laughing at you.
      • You also need a model for New Orlean's endemic corruption. What the approved plans specify and what gets built are two different things.

        In most places, there is always someone willing to cut corners if it will put extra money in their pocket.

    • Re:My inner Materials scientist just got shot. (Score:4, Informative)

      by Jim_Callahan ( 831353 ) on Sunday February 05, 2006 @04:39PM (#14647120)
      This is only true if your engineers are completely incompetent and mix the concrete incorrectly in places. Otherwise, concrete is much more regular at microscopic level than glass or plastic, which are both amorphous solids. And if your engineers are that incompetent, a good model for cracking is not going to save you. As to larger-scale irregularities, they're generally irrelevant if you know the structures of the component grains or regions.
       
      We already know the strengths of the materials we use to build things, because we test them beforehand. The reason new models are important is that they give us a better understanding of _why_ the strengths turn out the way they do.
       
      P.S. A +5 insightful attached to an assertion that a mostly crystalline solid is less ordered than an amorphous solid makes me cry inside. Everyone makes mistakes, but moderators aren't supposed to encourage it :-P

      • Re:My inner Materials scientist just got shot. (Score:5, Insightful)

        by theguyfromsaturn ( 802938 ) on Sunday February 05, 2006 @07:43PM (#14647654)

        Where the heck are you getting this idea that concrete is more regular than glass? You seem to be talking of a laboratory prepared CEMENT mixture. Yes, certainly if you prepare glass and cement in a laboratory condition, your hardened cement is likely to be more regular. There are, many considerations that makes this untrue for real life construction.

        The first thing to understand is that concrete is NOT cement. Concrete is a MIXTURE of cement and agregate. You can use all kinds of things for agregate, gravel and sand being the most common. Sometimes some fly ash from blast furnaces is added. Engineers normally use lower factors of safety for concrete than steel because the uncertainties are greater. When you test concrete to failure, sometimes the fractuers cut across the agregate grains where the cement bond was stronger than the agregate, other times it will follow a path around strong agregate particles.

        The other thing to know about concrete is that is it NOT made in a factory, under controlled industrial conditions and unit testing. Sure, you may get your concrete mixed at a concrete plant and the trucks, but eventually it has to get to the field. Then it must be placed... and the experience and professionalism of the workers is very variable. Furthermore, concrete needs to cure in place. The water content of the concrete during this stage is important since it needs water for the chemical reactions to harden the concrete. But then again there is an optimum value. The chemical reaction is also helped by high temperatures. So weather conditions and placement conditions will affect the final product.

        And of course, portland cement is a strong alkali. It can actually react with the agregates themselves which can build up stresses and cause cracks inside the concrete independently of external stresses. You may have witnessed this alkali-aggregate reactivity in concrete if you see cracks in concrete that seem to be humid, even what it hasn't been raining, and somtimes oozing a bit of white foam.

        In final analysis concrete is a highly nonuniform construction material.

        It can also added that most of your levees, and most likely the sections that failed, are probably earthworks. Therefore whatever the uniformity or lack thereof of the concrete, it would have done nothing for the leveees. Cracks are only meaningful in materials that need resist tensile or bending stresses. Needless to say, that is NOT how earthworks are desined.

    • It may mean something that they listed plastic, glass, and metal, and it may not. In fracture, these materials usually behave very differently. Glass is nearly a perfect brittle fracture, while the tip of the crack in plastics often distorts, so that it doesn't concentrate stress as effectively, meaning that it takes more energy to propogate the crack through the material.

      However, the article doesn't say anything about their conditions, and it's possible to construct samples of most materials that will frac
    • I smell someone whoring their ideas for grant money.

      Big bucket of Katrina Federal Cash! How can we get some?

      Ah ha! Stress fractures in metal are in some very vague sense similar to stress fractures in rock, and therefore dirt, and therefore levees and therefore profit!

      Free hotel, Free room service, Free Willy! WOOHOO
      • "I smell someone whoring their ideas for grant money."

        Ditto, magic formula with no details applied to a completely different (high profile) problem. BTW: Glaziers have been predicting how glass will crack for a very long time and have embeded the technology in a $5 tool called a "glass cutter".

        Using statistics to predict individual crack formation in a uniform material such as glass or metal is like predicting an avacado using the periodic table
    • Israeli physicists from the Weizmann Institute have used a new approach to study how materials break.

      Plaestinian bombers?

      -Eric

  • Interesting (Score:5, Interesting)

    by tsa ( 15680 ) on Sunday February 05, 2006 @03:35PM (#14646920) Homepage
    I'm currently involved in the making of 3D nano devices. One of the steps involved in the making of these things is the breaking of a silicon wafer. This is currently the least reliable step in our process, and we sure are very interested in ways to improve this.

  • Har (Score:1)

    by Quantam ( 870027 )
    I nominate this discovery for the title of highest ratio of actual importance to perceived importance.

  • Slightly misleading summary (Score:5, Insightful)

    by gunpowda ( 825571 ) on Sunday February 05, 2006 @03:38PM (#14646936)
    It would be great if submitters of content actually read it, and made it as 'brilliant' as their attempted irony.

    The news release even says that it could have help engineers predict 'exactly how much pressure the levees protecting New Orleans could withstand before giving way.'

    No, it doesn't. That's a rhetorical question in the first paragraph.

    • Not even a very good rhetorical question as the article refers to studying cracks in a given material. The levee failures in New Orleans were very probably not caused by a failure of a single element. Bending loads and the ability to resist them in systems made up of steel, concrete, and compacted earth are considerably more complex than propagation of a crack through a single material. Water movement and the undermining of foundations complicate thigs further by adding some dynamic loading to what are ge
    • They don't exactly carry the May issue of "Physical Review E" on the local newsstands, and the press release came out this month, so nobody knew about it... so no, it couldn't have helped to predict anything.

      Plus they already knew how much the levees could hold. Katrina exceeded that.

      (Yes, it weakened to a Cat 3 as it made landfall, but it was for all intents and purposes a Cat 4 storm as it approached, storm surge and all.)

    • Whoa, calm down. This is from Roland Piquepaille for christ's sake, not Woodward and Bernstein.

  • Quote from article (Score:3, Funny)

    by CrazyJim1 ( 809850 ) on Sunday February 05, 2006 @03:53PM (#14646983) Journal
    "Scientists all over the world have experimented with crack(ing)"
  • I've found that a good way to stop a crack from spreading is to drill a hole at the leading edge. Then use epoxy or whatever on the crack itself.

    Maybe a bit off topic, but could be useful.
    • actually, yes and no. the cracks they are talking about start out initially a few nanometers in length. They are also typically internal. BUT, there are many composites which include nodules of a glue-like substance and nodules of a catalyst that enables the glue to harden. SO the idea is, a crack in the material forms. When it propagates, it hits a glue nod, and breaks it open. the glue expands and fills the crack. The catalyst activates the glue and hardens it, effectively stopping the crack from propaga

    • Re:Best wat to stop a crack... (Score:4, Informative)

      by hairykrishna ( 740240 ) on Sunday February 05, 2006 @05:19PM (#14647235)
      Yep, that's because the raise in stress at the crack tip is proportional to it's radius. By drilling it out you increse the radius and decrease the stress.

  • Obligatory (Score:2, Funny)

    by wootest ( 694923 )
    These people crack me up.

  • Punny (Score:4, Informative)

    by h nu per lmda ( 952171 ) on Sunday February 05, 2006 @04:02PM (#14647021)
    I hope the title was meant to be punny, because a model on cracking is nothing new. There are currently many models that work for crack propagation in composites (of metal, polymers and amorphous materials). Every research groups CLAIMS that their model allows them to gain the best insight, because saying anything else:
    forfeits further research dollars.

    Until the model is explained in further detail and some source code is released, rather than the typical hand-waving, hype and money generating BS, this "breaking news" is nothing but hype.

    -PhD student. Metallurgical Sciences.
    • I'm glad I wasn't the only one that reacted this way to the article. Remember, though, that it's posted on the internet so that a bunch of people can go "oh, yeah, knowing how cracks propagate really _would_ be handy, i never would have thought of that" and to assure them that smart peoples be working on the problem. So to the intended audience, it probably _is_ new.
    • Until the model is explained in further detail and some source code is released, rather than the typical hand-waving, hype and money generating BS, this "breaking news" is nothing but hype.

      Try looking in Dr. Procaccia's list of recent and unpublished papers [weizmann.ac.il], although I don't know if this specific model in question is included in any of those papers.

    • I aggree.

      There are so many models on crack that this is hardly news at all. Predicting when a model on crack will turn into a crack ho is a fairly exact science and I don't believe that this article addresses at all.

  • "Blessed are the cracked, (Score:4, Funny)

    by chris_eineke ( 634570 ) on Sunday February 05, 2006 @04:06PM (#14647027) Homepage Journal
    ... for they let in the light." - Spike Milligan

  • Why is this on /.??? (Score:4, Insightful)

    by geneing ( 756949 ) on Sunday February 05, 2006 @04:16PM (#14647055)
    Dear Editors,

    Why is this article posted on /.? At best this is a report of a minor advance in a well established field. Hundreds of such advances are made in every field every week. Yes, PR department at Weitzmann Inst called it a breakthrough but that doesn't make it into one.

    Is it possible to limit the science postings to real science news? Maybe have editors who know the field evaluate the postings before hand.

    • In other news, PI has at least one more decimal place than what has previously been calculated.

    • Re:Why is this on /.??? (Score:3, Insightful)

      by Otter ( 3800 )
      Personally, I couldn't care less about how many stories Roland Piquepaille submits or whether they link to his blog. The problem is that all his science-related links are like this one -- some press release about a respectable but routine publication, selected seemingly at random and spun into a revolutionary new breakthrough.
    • Maybe have editors who know the field...

      That's a good one! Hahahaha....

    • Re:Why is this on /.??? (Score:3, Insightful)

      by wass ( 72082 )
      Uggh, didn't realize this was a Roland article.

      But anyway, are you that new here? Nearly all /. science articles report on press-releases like this, this specific article is no exception. In fact, that's the whole point of press releases, it's very rare to come across any 'earth-shattering' discoveries in the sciences these days. So on one hand it's cool that of the thousands of research projects going on making small but steady headway, a few of these results are reported here. For example, I certain

  • 2000 years too late, the veil split already.
  • This article was a disappointment - the science of fracture mechanics is at least a half century old, and quite well developed. It *is* a reasonably hard problem to predict crack growth, direction, speed etc in metals (say in an aeroplane structure) but there is software that'll do it, given a big computer, a good FE model, plenty of time, and lots of good material data (usually at least one of these is missing, in my experience...)

    However, my quick read of the aricle did not enlighten me at all as to
    • I've just read it again so I can answer my own question...

      They are scientists trying to work out the "why" and "how" at a microscopic level.

      Us engineers, OTOH, couldn't really care less, as long as we've got good enough tools to ensure our macroscopic things like aeroplanes don't break unexpectedly...

      ...and these tools don't always come from scientists, they often come first from engineers.
  • I thought this was going to be about tracking people who crack software.
  • It used to be the running joke that any scientific research would claim to possibly be leading to a cure for cancer. Now I guess the new paradigm is that it can be fixing the New Orleans problem instead.
  • They failed due to base erosion caused by overtopping and/or poor 'transition points' where embankments went from natural earthen mounds to man-made concrete slabs. Has there been something more authoritative (that included cracking) since the ASCE's report?

  • Crack dealers (Score:3, Funny)

    by RedLaggedTeut ( 216304 ) on Sunday February 05, 2006 @05:26PM (#14647259) Homepage Journal
    Well, I heard the rumor that the Israelis had taken over the dealing in XTC, but hey, now that I heard it confirmed on slashdot that they're now taking over the crack trade as well, my, gosh.
  • Gimme a break (pun intended):

    The news release even says that it could have help engineers predict 'exactly how much pressure the levees protecting New Orleans could withstand before giving way.'

    Come on. That's ridiculous. If only... Pffff!!! Besides the fact that it doesn't matter when those levees would have broken, because they would have anyway, this is nothing more than FUD to drum up money.

  • American Committee for the Weitzman Institute? (Score:5, Interesting)

    by Animats ( 122034 ) on Sunday February 05, 2006 @06:39PM (#14647482) Homepage
    This press release comes from the American Committee for the Weitzmann Institute of Science. [weizmann-usa.org] Who are they? The Weitzmann Institute is a good research school in Israel, but this "American Committee" thing is apparently some kind of a fund raising operation. The return address on the press release is from Janine Gordon Associates [jgordonassociates.com], "Where corporate and brand reputations are built, enhanced, and transformed." They also promote "bankrate.com" and Bridal Guide magazine.

    Janine Gordon Associates specializes in placing favorable PR pieces, rather than direct advertising. See their case histories [jgordonassociates.com] page, where they boast about how they plant stories. (Note: annoying all-Flash site.)

    This is a Roland the Plogger story, of course. But, for once, none of the links benefit his search engine ranking. So one wonders if Janine Gordon Associates uses Roland the Plogger.

  • This is great! Now if only they could help me keep tabs on my glass pipe, too. After the first couple of twenty rocks I always seem to lose track of the pipe and then I end up having to smoke out of an old lightbulb.
  • Don't build a goddamned city below sea level!

    Anti-US ramblings aside, I am fascinated by this very topic and have pondered upon the science behind destruction, especially with regards to tensile strength. Why do things smash when we hit them ? Why does wood split when you drive a screw/nail through it ? I think this research will not only give us very advanced insight for future building projects, but it might also trickle down to everyday uses like fasteners and glues, or maybe a better wood axe.

    The bes

Slashdot Top Deals

Whoever dies with the most toys wins.

Close