I went into a restaurant that was less than a year old and they had a very large laminated beam holding up the ceiling and it was twisted as badly as your CGI example and was only inches from sliding off the center support post. I called the manager/owner and discussed the impending disaster that the beam posed. He insisted that they had an engineer design it but I pointed out that if it had been PROPERLY engineered, it wouldn't be twisting, and it wasn't a matter of IF but WHEN the beam was going to collapse. I went back a couple of months later and the beam had been replaced.
Hi Gary, that's terrifying! Twisting can be very destructive to sections that are not designed to handle it (such as wide flange beams). In your case, a laminated beam (assuming rectangular) would be a bit better at resisting twisting loads though it might be problematic for the laminations. In any case, I would still not bet my life on any design that undergoes excessive deflections/deformations. The serviceability limit state (SLS) is also an important part of the design process. I would not trust any structural member that does not comply with it and judging by your description, that beam did not comply with it.
@@TheEngineeringHub - UPDATE: My curiosity was peaked and I went back to the restaurant. Apparently the replacement beam didn’t work out either. They removed the beam and the whole roof (it was a patio cover) and replaced it with an artistic shade installation.
Years ago, I had an engineer on staff that did the structural engineering and planned the steel elements in mostly wood framed structures. He designed almost exclusively using square tube. Now I understand why.
Closed sections are really good for edge members or other applications where asymmetric load cases may arise. That is the beauty of engineering, finding the best and cheapest solution for a particular problem. Thanks for the comment Roger, we love it when viewers contribute to the discussion!
@@TheLuminousOne No reason to throw insults! He is a very skilled engineer. The tube shapes integrated better with the mostly wood framed structures and the seismic retrofs that we were building at the time. He used open sections in commercial buildings that we built.
@@rogerhodges7656 just going to point at bones, and leg bones especially - change cross-section and internal density depending on the loads imposed from the joint ends and muscle attachments.
@@bbgun061 because any tube section is more difficult and expensive to design for the joints, there isn’t any standard sizes as big as the larger sizes in beams and going to a welded box section is expensive compared to a beam. Nothing wrong with using an I beam or column so long as they are correctly loaded.
@@BenDecko2023 I'm not exactly sure if I understand the premise of your questions, but I'll try... If by "used" you mean old repurposed lumber that was salvaged, then it can be much stronger than the lumber that is harvested today if it was harvested in an old growth forest, usually around 20%, but it can be as much as twice as strong. If it's just "old" scrap lumber from a previous job, then it might have the benefit of have less moisture in it, so it'll shrink less as the building settles in, but that would be a stretch and hard to quantify. For the second part, if you're talking about ripping repurposed lumber on a table saw to get it to a smaller dimensional size then that could be fine, but care has to be taken not to expose any knots or checks on the top or bottom surfaces of the lumber, as doing so would significantly decrease the strength of that piece. Having edges free of those defects is the primary way in which lumber is graded for strength.
@@elektrolyte 🙂👍and remember - "When members are compressed, they tend to deflect laterally". My ex is still suffering after ignoring this trenchant plea for caution.
Being a welder and working on many structures using beams you get the instinctual sense of how a beam is weaker in its torsional axis because they twist much easier then they bend
Words are not enough to thank you and appreciate what you are doing here. I had steel III in my Masters of civil engineering, and i passed the course with excellence but i swear god i never fully comprehend the meaning of LTD as much as i saw your video. Thanks again, wish you all the best
“Again you have heard that it was said to those of old, ‘You shall not swear falsely, but shall perform to the Lord what you have sworn.’ 34 But I say to you, Do not take an oath at all, either by heaven, for it is the throne of God, 35 or by the earth, for it is his footstool, or by Jerusalem, for it is the city of the great King. 36 And do not take an oath by your head, for you cannot make one hair white or black. 37 Let what you say be simply ‘Yes’ or ‘No’; anything more than this comes from evil.- JESUS Christ, Son of GOD and GOD HIMSELF
Swearing shouldn’t be a thing, the only reason why swearing is a thing is because of lies, and untrustworthyness, lies come from evil, and evil from the evil one. But JESUS has already one, GOD bless you
Not only are the beams slightly bent to start with the web is never in the center. I worked with beams in construction for many years and rarely got one that was centered.
Yeah, the manufacturing tolerances for hot-rolled products are pretty slack. It starts at bar stock. Unless you're dealing with cold rolled, calibrated or even machined or ground material, you could be dealing with variations that are visible to the naked eye. Some hollow stock (square round etc) is also hot-drawn through dies, where the inside die is held in the center only by an equilibrium of material pressure from all sides, so the material thickness can vary by a few percent all around. Similar thing for aluminum extrusions.
Those tolerances are considered when determining the design capacities used in building codes. If you look at the raw test data for ultimate failures of W-shapes, they can vary well over 30%, which is part of the reason we wind up with a safety factor of more than 3 when all is said and done. Obviously that doesn't help the people who have to put them all together LOL!
Excellent presentation. Edit: top loading the beam may have resulted in even less weight required for failure, as the blocks and load placement actually “strengthened” the beam.
My father was a foreman for a construction company. Now i understand why my father chooses rectangular open beam for our home to support the second floor (wooden and pest with ermite).
Visit brilliant.org/TheEngineeringHub/ to get started learning STEM for free, and the first 200 people will get 20% off their annual premium subscription.
If you see the top member as a compressed bar, you get the buckling. Put the buckling member into the beam and you have how the twisting is generated. For the buckling cause: it's not just the uneven manufacturing, it's also the uneven installation and loading.
Its not the beam design thats faulty, its the installation. The #A or #B type steel decking on top of the beam should have fasteners that attach it to the beam that prevent what you are pointing out. That type of beam isnt designed to be stand alone. There are thousands upon thousands of buildings designed with this type of beams that are perfectly fine. Pointing out a few instances of improper use doesnt mean its engineered incorrectly.
I-beams are the best for static axial loads, when the load becomes dynamic and/or torsion is introduced it can become complicated to calculate the expected deformation and most likely point of failure (for non engineers.) Defects in material, high point loads, different welding procedures & varying degrees of fit up can further exacerbate these uncertainties. To actually attain full load capacity of an I-beam would require bracing, gussets, and on longer span tying in the top & bottom flanges which in many applications does not happen. Coming from a guy w/ a fabrication background, I always err on the side on building it heavier w/ additional support, bracing, gussets, and extra contact points. I appreciate your videos as theres a huge gap between a fabricator & the engineer that draws up the plans and I fully intend on closing this gap. Best regards.
LOL I know a lot of engineers who struggle with those calculations to... Though it's not as bad as it may seem; these experiments used sections that would be considered extremely slender, leading to elastic and catastrophic failure from local flange buckling and web crippling. In reality most construction avoids these sorts of members except in special cases such as on bridges, roof girders in large warehouse style buildings, ect. And in those cases, out of plane bracing is a primary consideration. Most sections commonly used are relatively compact (think short and stubby with think web and flanges) and have similar design values whether they are assumed fully braced, or completely unbraced, so even if an engineer has no clue what they're doing, it's unlikely to result in catastrophic failure.
They are super flexing in the weak axis. And must be restrained to stop localized buckling. Usually they are tied into the concrete deck with shear studs. Or at least welded to the metal decking.
In the container shipping business, the Container Chassis uses a nominal 12" rolled I Beam. Those chassis are subjected to forces described in this article when they are not matched to a Container, repairing those Chassis can be difficult.
i would not call it 'weakness', rather a section property which must be adequately addressed in design. after reading all the comments below, there seems to be some confusion regarding engineer's choice of member section for any particular application. the engineer will check member properties in his steel manual (or the NDS if he is working with wood) and design appropriately for the application. video demonstration of lateral torsional buckling is cool - thanks for posting! one comment below that all lateral deflection is due to eccentricities in the member isn't necessarily true, although eccentricities will contribute to that effect. neither can residual internal stresses within the member be ignored. pilebutt story: a brand new freshly hot-rolled wide flange set to cool half in shade and half in direct sunshine was shown to have significant internal stresses beyond manufacturing tolerances and was rejected! also check for repairs/modifications to the beam, as cutting/welding can induce internal stresses. compression flange and tension flange are connected by the web, which is generally thinner than the flanges and connects the flanges, forming the force couple. the transition from flange to web is where the interesting things start to happen. you might consider make a video about web crippling. Registered Civil Engineer - retired bridge builder - i cant tell you how many wide flange beams i have installed and removed - clamped & braced - inspected for damage and authorized for continued use..... apply the FS!
very clear and intuitive sequences. going from the actual beam for visual reference - real world - back to line graphics illustrating forces and axis. this is how I learn and visualize regarding solutions reached on my own. works well. thank you
Great breakdown. What I can't wrap my head around is this. Open beams are used, largely because they are the least material for a given strength. And when they are inherently constrained from torsional flex they are great. As soon as you put them in a scenario where special anti-torsion construction is required in addition to everything else, I don't understand why it's not common practice at that point to just use a box beam instead of some complex x-bracing, etc. Any structural engineers out there care to weigh in?
I’m a mechanical engineer but work with structural engineers. I agree that with the necessary and essential cross bracing, it seems like there is a point where a closed section becomes a better choice. I design undersea systems and it’s common to avoid closed sections because we can’t visually inspect the interior surfaces during service. Also, it’s difficult to coat and verify coatings after manufacture. I doubt these are the reason that structural engineers prefer wide flange beams, but just thought I’d mention reasons from my perspective.
@@danielhoward755 You may very well be onto something there, and I had the same thought after the fact when thinking about bridges and such, that's almost certainly a huge factor. For other cases where beams aren't as exposed to the elements, it's likely just a matter of using what's commonly available.
If you ever want a good example of torsional weakness in I beams, there are a few videos where bridge beams are being transported as oversize loads and the dolly wheels at the back have an issue and the beam falls. the back of the beam will lay over and be napping on it's side, and you can watch the load transfer up the beam until...it flips the cab of the truck on it's side after a noticeable pause. even over 100', you can have a 90* offset to your ends without much work, yet the same beam is literally holding up bridges
Great visual presentation. I am an EE, but love working on cars so I find myself wading into ME territory often. I see this in pickup track frame design. Properly braced C-channel frames work well, but a fully boxed frame will be stronger still. It's interesting to look at the existing designs and how the cross sections change or where the bracing is applied and try to work out the forces in the steel. When modifying something structural like the frame it is useful to understand why something had been originally engineered a certain way.
Yup. I am a EE to but I used to haul heavy machine tools. An old welder introduced me to Boxing. Boxing is when you weld steel into the truck frame to box the strategic parts of the frame. Hot rodders often boxed their frames for strength. I always used to design beams initially by looking at the sectional modulus and ultimately the moment of inertia. In timber framing you don’t always deal with the slenderness ratio of thinner members. But slenderness ratio and ultimately beam stability do play a significant role. Even in post compression, you need to check your buckling modes as much as your initial compression loads. Buckling will ruin your day and LTD is one of the worst for sure.
Vehicles are also designed to respond properly to certain kinds of crash scenarios, so open v closed beam may be an actual designed in decision to make sure the engine goes under the car in a front end collision, and similar things.🙂
@@noyb7920 I was a volunteer tech working on coffin nosed Stanley steam cars. Wood bodies, no seat belts and a boiler large enough to heat a house! Worked on muscle cars in my youth including 426 hemi challengers. Modern vehicles are an over priced joke! Loaded full of obsolete consumer grade electronics and plastic. Designed to be snapped together like lego blocks and impossible to work on. You need an arsenal of snap on tools to get anywhere including stubbies, knuckles, imperial, metric, 6 point and 12 point shorties and extended sizes etc. You have to take half the car apart to fix anything. They are designed to go 200K and then off to the scrap yard. Now they have complex touch screens. Really People! If driving with a cell phone is dangerous, what about farting with a touch screen to get your iTunes account to play the latest hits at 90 MPH in congested freeway traffic. Stupid! When they crash, they virtually fly apart in a crumpled mess. Even minor accidents are total losses! Now the insurance companies have to cover this joke so your insurance rates are thru the roof! The marketing folks put lip stick on this pig and get everyone to buy these plastic buggies on seven year contracts. The cars are done at the end of seven years so you get to buy another one. Then they now require you to use fluids with secret recipes and huge price tags. What happened to old fashioned lubricants and coolants? And now you can’t get a manual tranny! All computerized automatics! An automatic is a machinist’s wet dream! And they are at least twice or more expensive than an old gear box. Have we become so lazy that gear jamming is such an inconvenience? I am glad to be facing retirement as I don’t want any part of this nonsense!
@@devmeistersuperprecision4155 Um, I didn't subscribe to this newsletter, and modern cars are not that hard to work on. See... most any car repair channel on youtube.
Part of the issue here is the length of the beam and the Gauge of the steel. A long piece of thin steel in any configuration will be susceptible to torsion and buckle.
Absolutely, the length has a higher order power in the denominator, which makes the whole equation of torsional buckling basically governed by the length of the beam
Amazing content, i hope that our engineering proffesors teached us structural mechanincs like the way you do. Can i know what visualization program you used to modell the lateral tortional buckling of the steel beam in 4:58?
The picture at 1:00 is worth a thousand words. When the beam is under load, both the top plates and bottom plates are *trying* to stretch, with the bottom wanting to stretch more. But, they _REALLY don't want to stretch!_ With the strap in the middle holding them together and keeping them from stretching different distances, there is only one way for them to go that keeps them from trying to stretch, so each can maintain as close to their original lengths as possible, which IS _the same length for each._ "Come on Baby, let's do the Twist!" This video explains why Engineered Trusses in floors are WELL cross-braced, nowadays.
I have floor trusses there o.c.16" 2x4 constructed and 18" tall they span 28'. I have squash blocks around the rim with a 1.25 osb sub floor. So far so Good.
Your loading is concentrated on very focused points along the beam. The load needs to be evenly distributed which helps the beam to function as it was intended.
Did he use the weakest looking beam he could find to make his point? Just look at that shitty ass beam, looks like you could break it by jumping on it.
@@lylemcdermott2566yes he did. He wanted to demonstrate failure safely. If he used a beam that would break at one ton it would take longer to load and failure could have caused damage to the person doing the loading. He got his point across and wasted less material with his demonstration.
@@lightningdemolition1964 but he drilled holes through the beam, didn't load it like it should have been loaded ( on the top) and beams aren't working that way in the real life scenario. For an engineer it's crap work.
@@lylemcdermott2566 Not to mention he didn't use H beams. He bolted 2 U channels together, totally weakening it AND introducing weaknesses that wouldn't be there with a proper H beam. Did the same with the "box" beam, again 2 U channels bolted together to "create" a box beam.
Thank you changw 🙏Reference [3], [4], and [5] are absolute gems. Luckily, as engineers, we have studied these books for many years which makes it a bit easier to provide citations but still, the animations and/or experiments do consume a lot of time.
Used and wore out a wide flange beam In A wood spliter and was planning on doubling thickness of top flange where the spliter knife rides. Your thoughts please on this? Thankyou
I think this is one reason we newer used open beam framework in the (north of) Nord for wooden buildings in historical times. Whan it was used in farmers buildings the beams are always square, never rectangular (in modern sense) in historical buildings here. Open beams where never used in living houses here, the roof on my old farmhouse is cladded with 2" plankings on both sides, and i think it is square beams under it in order to withstand the weight of wet snow.
That is why composite I-beams, and beams without fillets between the body and the flange are better. Because well filleted i-beams that are do not have wide flanges are far stronger across multiple axis. Neither of those beams are typical of properly selected material thickness. These are very good examples of silly minimalism in materials selection. The only thing one need do during design is to choose a failure mode that suits your needs. Because here is the thing, your test setup also failed to properly secure the ends of the beams you chose, so failure was inevitable because the beam never transmits tension to the points of support.
The problem seems much exacerbated by the excessive slenderness of the truss and the lack proper constrains at the extremities. This allows the upper wing if the truss to become unstable, due to lack of axis tension, in a similar way as a free slender plank become unstable under vertical load faster then one which has working constrains. What you are doing adding flanges to the side of the section is increasing the area of of it, and so reducing the slenderness of the truss itself. I guess standard truss dimensioning, like the IPE system suffers less of the problem because it accounts for the issue. I also imagine Baloon Framing is most vulnerable to this kind of problems
So a crystalline and straight beam would not experience this failure? (Due to having no imperfections even down the the molecular level, assuming the molecular bonding angle and resulting crystal structure are not introducing a mechanical flaw)
At 7:25 you say that the root cause is 'the imperfect manufacturing and construction process'. While that's part of it, I think it would be better described as the design being inherently unstable at high loadings. Vertical load on a deflected beam causes the beam to deflect further. The initial deflection could come from manufacturing or installation, but even with a perfect beam installed perfectly, there are all kinds of things that could lead to deflection. Wind loading, earthquake loading, even just uneven roof loading (the next beam over has an AC unit sitting on it, this one doesn't, so it sags a few mm less and is picking up more load from one side) are all things that could push a beam off the perfectly balanced knife edge.
Absolutely agree, even a beam that's perfect down to the last molecule would experience this failure because any external actions (a flap of a butterfly's wings) would cause it to buckle.
Seems like if two beams of equivalent material, one closed and one I beam, fail at such wildly different point loads when installed the same way; the use of open beams is only justifiable if there is a specific use for the flange or if the difference in manufacturing cost is really so great... but except at very large sizes it seems that both are priced pretty similarly.
The sections bend in the tranversal plane and generate horizontal loads, contrary to what is hypothesized in the calculations, hence the torsional buckling?
The calculations for the in-plane bending still hold. The only problem is that the torsional buckling happens first i.e. for a lower load (in slender sections) therefore the bending capacity does not govern. The torsional buckling happens due to the fact that the top of the beam is being compressed. If you imagine the compressed half of the beam as its own column then it's easy to see that it will want to "kick-out" of place. Opposite to that, the stretched part of the beam wants to stay where it is causing a twisting motion in the beam. I wouldn't say the calculations are wrong it's just a different failure mechanism.
You did a fine job of explaining, but perhaps an explanation of the mathematics for the mathematically challenged. I would love to try and understand math more in-depth.
The math shown is nothing more than empirical values and equations, meaning they are simple formula that is contrived to match experimental test data, it's not intuitive at all and trust me, seeing it would not lead to any significant insight. I'll post a link below of a video that shows how beam deflection is derived based on simple geometry which shows what I think you're asking to see here. That sort of elegant mathematical derivation is not how the equations shown in this video work. ruclips.net/video/g8luMxP5Yo4/видео.html
Interesting. I wonder if these equations could be made generic, if they aren't already, and have no bias for a particular metal so that the cross-section of a metal subject would be described and performed upon, analyzed and apply to all metals, if not all materials and substances, so there were no flaws in the equations and the same equation could be used for all cross-section shapes and materials. Thanks for the video.
Even though they are not used in construction, I’d like to see how a cross shaped (like a X but verticale and horizontal instead of diagonal) cross section would hold up compared to the I beam shape or the box shape. Intuitively I think having the wider width at the neutral point between tension and compression would add strength, but maybe not.
It would be substantially weaker than an equal weight box beam. A beam's resistance to bending is modeled using, among other properties, the 2nd moment of area of the beam's cross section. The result is that areas of a beam further from the neutral axis add more to the strength of the beam than area close to the neutral axis. This is why I-beams are so common: they're an extremely efficient shape for a strong but light beam.
@@markjacobson4248 but wouldn’t the extra width at the neutral point add more resistance to the twisting as the I beam is loaded? If the load is placed centered on that neutral point it would be pulling the horizontal part down, along its weak axis, but also along the vertical part’s strong axis. As demonstrated, that load would cause the vertical part to deflect sideways, but the horizontal part would counter that by having a greater width at the neutral point. I’m not an engineer, or even a student of engineering, so please don’t think my question of comment is stupid. I’m genuinely interested in this and am learning from it.
@@Linusgump As a simplified rule, the material in a beam that has the greatest contribution to the beam's resistance to deformation is the material that would experience the greatest local deformation to accomodate the overall deformation. Any material on the neutral axis of any deformation, whether twisting or bending, contributes nothing to strength because it doesn't deform at all. It's definitely not as straightforward with a complex motion like that in this video, where you have twisting around one axis and bending along two others, but the general rule still holds. A box beam would do a far better job of resisting deflection per weight. In general, the further material is from the neutral axis of any deformation, the more it matters. I-beams and H-beams are common design to maximize the material away from the neutral axis in the top and bottom flange, while the web between the flanges actually has minimal contribution to the bending strength of the beam. A hollow tube is a common design to maximize twisting resistance of a beam or structural member, as that maximizes the material away from the neutral axis of twisting, the central axis of the member. A box beam is a good compromise between both those designs, and is good for situations like this.
Even though I'm not an engineer, I could see right away that attaching stuff to an I-beam like in the thumbnail is a really bad idea. Even without calculations, the vertical part of the I-beam is clearly too tall and thin for that, and the horizontal parts are too thin as well, and so the slightest load can lacause deformations, which will eventually escalate to much serious buckling and failing. Either one redesigns the attachments or use different shaped (square sectioned) beams.
You see these on the highways every day. Flatbed step decks and heavy haul lowboy trailers. It's a ladder beam style construction with a crown in the center to combat bowing. This is why the run a spread axle and have bridge laws.
3:20 Now, THAT’s stupid scary. The smallest slip will cause the beam to slide out and chase someone. The yellow grips should be placed considerably wider. You can see it even start to wobble because the fulcrum is so small.
I am an EE but know now why structural engineers are brought in to do the heavy calcs for stuff like this now - for CPD it's brilliant as I never liked mech eng much - thanks 😁
The torsional stiffness of an open beam is 100s of times smaller than a closed rectangular cross section beam of same dimensions, material and thickness. This means that stresses on open beams must be strictly longitudinal. Any other direction favors the closed beam, because it might be 25% heavier, but it is 100s times stronger.
I would also point out one little mistake, that the most stressed part of the I-beam at the time stamp ruclips.net/video/XSl7ZntK94E/видео.html would be on the other side of the web, because the compressive stresses from torsion and bending about both axes will sum right at that point. Besides that, very good educational video :) keep up the good work.
The critical most part of the beam is at the root of Junction of Web- flange in compression i.e. (tf+ Root radii) from the extreme end of the compression zone.
6:58 hi, must the member that restrain the compression side of I beam be diagonal member? Can a horizontal member ties to the compression face of I beam restrains it from moving laterally?
Horizontal braces are also fine. You can see this often in the roofs of Walmart, Costco, Superstore, etc. There are parallel trusses supported by larger girders. Usually, at the columns the truss is extended to the lower flange of the girder to peovide lateral stability. This is because there are large negative moments there, causing the girder's bottom flange to kick out.
Had to unload a 50 foot I beam from a box trailer, 2 foot web. It took the better part of the morning to do it without twisting it and ruining the beam.
Isn't the instability of the beam better explained by the center of gravity of the load being offset to any degree from the vertical element of the beam? A box section achieves better stability by having a wider distance between vertical members, not from having fewer imperfections.
Sure, but that's only partly the reason. The susceptibility to lateral torsional buckling (LTB) is also due to the poor torsional resistance of the I-beam. Closed sections on the other hand are much better at resisting torsion and therefore have higher LTB resistance (not necessarily because of the center of gravity). A box beam with a slit along one side (open box section) would geometrically be very similar to a closed box beam but would not perform as well in LTB due to the fact of not being closed resulting in weak torsional resistance. The weak moment capacity in the lateral direction of the I-beam is a reason more why it is weaker than the box beam. Manufacturing imperfections are important but the essence of the weakness basically comes down to "it's just the way physics works". In retrospective, maybe in the video we made it sound like the manufacturing is the weak link when in fact it's the laws of universe 😅
Usually the people designing a structure will know if there will be any loading that would cause a beam to fail, you should already know this. They would design in crossbeams to ensure the beam wouldn't flex, I know this and I am a retired electrician.
With regards to lateral torsion, to say that the root cause is material imperfection is a red herring. Material had better have been inspected prior to installation, beams had better be tested before they are used in production. Especially per findings on this disaster where a beam was not inspected or tested: en.wikipedia.org/wiki/Hyatt_Regency_walkway_collapse Material is always imperfect, that is its nature, account for it, assume it to exist. Just like in electrical engineering, where the tiniest semiconductor imperfection in a balanced but unstable driven circuit causes oscillation, often as intended. If you intend it to be rigid, make it stiff in all directions. Never engineer in 2 dimensions because in 3-4 dimensions it will fail. ~One Electrical Engineer
Hi not amouse, this imperfection is so microscopic that no inspection can detect it and more importantly even if it could, it cannot be eliminated. Instead it is just assumed that it is there. It is the nature of our world, nothing is perfect. The problem is that this microscopic deflection gets amplified when an unbraced beam is loaded and results in lateral buckling. If in a parallel universe we could produce a perfect beam down to the Planck length it would still buckle laterally because the load is not perfectly centered and will always prefer one side over another. If somehow the load was also perfectly centered then the beam won't buckle laterally but it would take the flap of a butterfly's wings to initiate this catastrophic process. So in essence lateral torsional buckling does come down to imperfections and natural preference of the material to one side over the other. Engineer's prevent this by totally avoiding it. No design ever allows for lateral torsional buckling to occur in normal operating conditions. In regards to your link, the Hyatt Regency walkway collapse is a very famous structural failure and we have studied and discussed it numerous times during our education and practice. A more rigorous inspection process and better communication most likely would have caught the error since it is absurdly obvious. But this is very different from the microscopic imperfections discussed in this video.
It is a great word 😅 I actually just searched the script and found only 3 matches for words starting with "intu." I'm not sure how many mentions are in the video, but in general, the voiceover follows the script. If it takes 3 drinks to get you hammered out of your head, then you can make a drinking game out of basically anything 😅
@christianlibertarian5488 they are extremely strong when they can be braced or when the top of the beam is connected to a supporting slab, which prevents it from moving.
As a fabricator I just make what the engineers say But now I’m need to know why use i beam at all if a box beam for the same weight is stronger and harder to work with and more expensive
Hi Aussie, a box beam is only stronger for unrastrained applications. For most applications , W-sections can be laterally braced to prevent lateral-torsional buckling. In that case (restrained), the w-beam (or I-beam) is significantly more efficient at carrying bending loads.
Your comparison test was NOT equal. The I beam has just the single thickness on the horizontal sections with ZERO fasteners strengthening them. While the box section has the horizontal sections doubled up with many fasteners strengthening it. It was not clear how the fasteners on the vertical sections compared. If they were equal, it made the box beam just that much stronger by the fact it had more material. The fact that the material is added where it strengthens the beam the most contribute to an unequal, and thus unfair test. Spot welds would add no material, but would be just about impossible to do on the box beam. Regardless, it would still do more to strengthen the box beam than it does the I beam. A fairer test would be I beams and box beams of the same height, width and the same weight. Made that way, not assembled from C sections.
![Twenty One Pilots - The Line (from Arcane Season 2) [Official Audio]](http://i.ytimg.com/vi/XCz7WUotXs8/mqdefault.jpg)
I went into a restaurant that was less than a year old and they had a very large laminated beam holding up the ceiling and it was twisted as badly as your CGI example and was only inches from sliding off the center support post. I called the manager/owner and discussed the impending disaster that the beam posed. He insisted that they had an engineer design it but I pointed out that if it had been PROPERLY engineered, it wouldn't be twisting, and it wasn't a matter of IF but WHEN the beam was going to collapse.
I went back a couple of months later and the beam had been replaced.
Hi Gary, that's terrifying! Twisting can be very destructive to sections that are not designed to handle it (such as wide flange beams). In your case, a laminated beam (assuming rectangular) would be a bit better at resisting twisting loads though it might be problematic for the laminations. In any case, I would still not bet my life on any design that undergoes excessive deflections/deformations. The serviceability limit state (SLS) is also an important part of the design process. I would not trust any structural member that does not comply with it and judging by your description, that beam did not comply with it.
@@TheEngineeringHub - UPDATE: My curiosity was peaked and I went back to the restaurant. Apparently the replacement beam didn’t work out either. They removed the beam and the whole roof (it was a patio cover) and replaced it with an artistic shade installation.
Compliments on your insistence to the manager! You probably saved Many lives! Thank you 🙏!
Clearly their insurance agent did not see that - if he had he should have refused coverage until repairs complete.
@@garyreed2206 good on the manager for listening.
Years ago, I had an engineer on staff that did the structural engineering and planned the steel elements in mostly wood framed structures. He designed almost exclusively using square tube. Now I understand why.
Closed sections are really good for edge members or other applications where asymmetric load cases may arise. That is the beauty of engineering, finding the best and cheapest solution for a particular problem. Thanks for the comment Roger, we love it when viewers contribute to the discussion!
@@TheLuminousOne No reason to throw insults! He is a very skilled engineer. The tube shapes integrated better with the mostly wood framed structures and the seismic retrofs that we were building at the time. He used open sections in commercial buildings that we built.
@@rogerhodges7656 just going to point at bones, and leg bones especially - change cross-section and internal density depending on the loads imposed from the joint ends and muscle attachments.
If the same amount of material in a box section is stronger, then why are I-beams ever used?
@@bbgun061 because any tube section is more difficult and expensive to design for the joints, there isn’t any standard sizes as big as the larger sizes in beams and going to a welded box section is expensive compared to a beam. Nothing wrong with using an I beam or column so long as they are correctly loaded.
Wow, I have never seen such a succinct and comprehensive review of LTB! And with an at-home experiment too! Great work. Thank you.
This is why blocking is extensively used in wood framing. It prevents torsional buckling that can happen even on lumber.
They Dont call them Squash blocks for nothing.
@@BenDecko2023 He said it, extensive use of blocking. Any weaknesses are rendered moot by the amount of redundant reinforcement.
@@BenDecko2023 I'm not exactly sure if I understand the premise of your questions, but I'll try... If by "used" you mean old repurposed lumber that was salvaged, then it can be much stronger than the lumber that is harvested today if it was harvested in an old growth forest, usually around 20%, but it can be as much as twice as strong. If it's just "old" scrap lumber from a previous job, then it might have the benefit of have less moisture in it, so it'll shrink less as the building settles in, but that would be a stretch and hard to quantify. For the second part, if you're talking about ripping repurposed lumber on a table saw to get it to a smaller dimensional size then that could be fine, but care has to be taken not to expose any knots or checks on the top or bottom surfaces of the lumber, as doing so would significantly decrease the strength of that piece. Having edges free of those defects is the primary way in which lumber is graded for strength.
Any video with the words 'member' and 'flange' in the first minute, is worth my like and subscribe!
LEGEND!!😍
@@elektrolyte 🙂👍and remember - "When members are compressed, they tend to deflect laterally". My ex is still suffering after ignoring this trenchant plea for caution.
Being a welder and working on many structures using beams you get the instinctual sense of how a beam is weaker in its torsional axis because they twist much easier then they bend
Words are not enough to thank you and appreciate what you are doing here.
I had steel III in my Masters of civil engineering, and i passed the course with excellence but i swear god i never fully comprehend the meaning of LTD as much as i saw your video.
Thanks again, wish you all the best
“Again you have heard that it was said to those of old, ‘You shall not swear falsely, but shall perform to the Lord what you have sworn.’ 34 But I say to you, Do not take an oath at all, either by heaven, for it is the throne of God, 35 or by the earth, for it is his footstool, or by Jerusalem, for it is the city of the great King. 36 And do not take an oath by your head, for you cannot make one hair white or black. 37 Let what you say be simply ‘Yes’ or ‘No’; anything more than this comes from evil.- JESUS Christ, Son of GOD and GOD HIMSELF
Swearing shouldn’t be a thing, the only reason why swearing is a thing is because of lies, and untrustworthyness, lies come from evil, and evil from the evil one. But JESUS has already one, GOD bless you
Not only are the beams slightly bent to start with the web is never in the center. I worked with beams in construction for many years and rarely got one that was centered.
Great point! Thanks for bringing this up!
Yeah, the manufacturing tolerances for hot-rolled products are pretty slack. It starts at bar stock. Unless you're dealing with cold rolled, calibrated or even machined or ground material, you could be dealing with variations that are visible to the naked eye. Some hollow stock (square round etc) is also hot-drawn through dies, where the inside die is held in the center only by an equilibrium of material pressure from all sides, so the material thickness can vary by a few percent all around. Similar thing for aluminum extrusions.
Yeah, the variations in the web can be a serious pain when the prints assume it's exactly centered.
Those tolerances are considered when determining the design capacities used in building codes. If you look at the raw test data for ultimate failures of W-shapes, they can vary well over 30%, which is part of the reason we wind up with a safety factor of more than 3 when all is said and done. Obviously that doesn't help the people who have to put them all together LOL!
Excellent presentation.
Edit: top loading the beam may have resulted in even less weight required for failure, as the blocks and load placement actually “strengthened” the beam.
My father was a foreman for a construction company. Now i understand why my father chooses rectangular open beam for our home to support the second floor (wooden and pest with ermite).
Visit brilliant.org/TheEngineeringHub/ to get started learning STEM for free, and the first 200 people will get 20% off their annual premium subscription.
If you see the top member as a compressed bar, you get the buckling. Put the buckling member into the beam and you have how the twisting is generated.
For the buckling cause: it's not just the uneven manufacturing, it's also the uneven installation and loading.
Agreed! Also, a nice short summary of the mechanism 👌
Its not the beam design thats faulty, its the installation. The #A or #B type steel decking on top of the beam should have fasteners that attach it to the beam that prevent what you are pointing out. That type of beam isnt designed to be stand alone. There are thousands upon thousands of buildings designed with this type of beams that are perfectly fine. Pointing out a few instances of improper use doesnt mean its engineered incorrectly.
I-beams are the best for static axial loads, when the load becomes dynamic and/or torsion is introduced it can become complicated to calculate the expected deformation and most likely point of failure (for non engineers.) Defects in material, high point loads, different welding procedures & varying degrees of fit up can further exacerbate these uncertainties. To actually attain full load capacity of an I-beam would require bracing, gussets, and on longer span tying in the top & bottom flanges which in many applications does not happen. Coming from a guy w/ a fabrication background, I always err on the side on building it heavier w/ additional support, bracing, gussets, and extra contact points. I appreciate your videos as theres a huge gap between a fabricator & the engineer that draws up the plans and I fully intend on closing this gap. Best regards.
Right On. Build it to be Safe and Last the test of time.
LOL I know a lot of engineers who struggle with those calculations to... Though it's not as bad as it may seem; these experiments used sections that would be considered extremely slender, leading to elastic and catastrophic failure from local flange buckling and web crippling. In reality most construction avoids these sorts of members except in special cases such as on bridges, roof girders in large warehouse style buildings, ect. And in those cases, out of plane bracing is a primary consideration. Most sections commonly used are relatively compact (think short and stubby with think web and flanges) and have similar design values whether they are assumed fully braced, or completely unbraced, so even if an engineer has no clue what they're doing, it's unlikely to result in catastrophic failure.
They are super flexing in the weak axis. And must be restrained to stop localized buckling. Usually they are tied into the concrete deck with shear studs. Or at least welded to the metal decking.
In the container shipping business, the Container Chassis uses a nominal 12" rolled I Beam. Those chassis are subjected to forces described in this article when they are not matched to a Container, repairing those Chassis can be difficult.
i would not call it 'weakness', rather a section property which must be adequately addressed in design.
after reading all the comments below, there seems to be some confusion regarding engineer's choice of member section for any particular application. the engineer will check member properties in his steel manual (or the NDS if he is working with wood) and design appropriately for the application. video demonstration of lateral torsional buckling is cool - thanks for posting!
one comment below that all lateral deflection is due to eccentricities in the member isn't necessarily true, although eccentricities will contribute to that effect. neither can residual internal stresses within the member be ignored. pilebutt story: a brand new freshly hot-rolled wide flange set to cool half in shade and half in direct sunshine was shown to have significant internal stresses beyond manufacturing tolerances and was rejected! also check for repairs/modifications to the beam, as cutting/welding can induce internal stresses.
compression flange and tension flange are connected by the web, which is generally thinner than the flanges and connects the flanges, forming the force couple. the transition from flange to web is where the interesting things start to happen. you might consider make a video about web crippling.
Registered Civil Engineer - retired bridge builder - i cant tell you how many wide flange beams i have installed and removed - clamped & braced - inspected for damage and authorized for continued use.....
apply the FS!
very clear and intuitive sequences. going from the actual beam for visual reference - real world - back to line graphics illustrating forces and axis. this is how I learn and visualize regarding solutions reached on my own. works well. thank you
Great breakdown. What I can't wrap my head around is this. Open beams are used, largely because they are the least material for a given strength. And when they are inherently constrained from torsional flex they are great. As soon as you put them in a scenario where special anti-torsion construction is required in addition to everything else, I don't understand why it's not common practice at that point to just use a box beam instead of some complex x-bracing, etc.
Any structural engineers out there care to weigh in?
I’m a mechanical engineer but work with structural engineers. I agree that with the necessary and essential cross bracing, it seems like there is a point where a closed section becomes a better choice. I design undersea systems and it’s common to avoid closed sections because we can’t visually inspect the interior surfaces during service. Also, it’s difficult to coat and verify coatings after manufacture. I doubt these are the reason that structural engineers prefer wide flange beams, but just thought I’d mention reasons from my perspective.
@@danielhoward755 You may very well be onto something there, and I had the same thought after the fact when thinking about bridges and such, that's almost certainly a huge factor.
For other cases where beams aren't as exposed to the elements, it's likely just a matter of using what's commonly available.
Great content! Just found your channel and these videos are extremely polished. Good work, hope to see more.
Welcome aboard! More content it's one its way!
If you ever want a good example of torsional weakness in I beams, there are a few videos where bridge beams are being transported as oversize loads and the dolly wheels at the back have an issue and the beam falls. the back of the beam will lay over and be napping on it's side, and you can watch the load transfer up the beam until...it flips the cab of the truck on it's side after a noticeable pause. even over 100', you can have a 90* offset to your ends without much work, yet the same beam is literally holding up bridges
As a musician I can't wait to apply this knowledge in my daily basis
Great visual presentation. I am an EE, but love working on cars so I find myself wading into ME territory often.
I see this in pickup track frame design. Properly braced C-channel frames work well, but a fully boxed frame will be stronger still. It's interesting to look at the existing designs and how the cross sections change or where the bracing is applied and try to work out the forces in the steel. When modifying something structural like the frame it is useful to understand why something had been originally engineered a certain way.
Yup. I am a EE to but I used to haul heavy machine tools. An old welder introduced me to Boxing. Boxing is when you weld steel into the truck frame to box the strategic parts of the frame. Hot rodders often boxed their frames for strength. I always used to design beams initially by looking at the sectional modulus and ultimately the moment of inertia. In timber framing you don’t always deal with the slenderness ratio of thinner members. But slenderness ratio and ultimately beam stability do play a significant role. Even in post compression, you need to check your buckling modes as much as your initial compression loads. Buckling will ruin your day and LTD is one of the worst for sure.
@@devmeistersuperprecision4155 Exactly right. I've seen boxing used on pick up truck frames when modifying them for off road use.
Vehicles are also designed to respond properly to certain kinds of crash scenarios, so open v closed beam may be an actual designed in decision to make sure the engine goes under the car in a front end collision, and similar things.🙂
@@noyb7920 I was a volunteer tech working on coffin nosed Stanley steam cars. Wood bodies, no seat belts and a boiler large enough to heat a house! Worked on muscle cars in my youth including 426 hemi challengers. Modern vehicles are an over priced joke! Loaded full of obsolete consumer grade electronics and plastic. Designed to be snapped together like lego blocks and impossible to work on. You need an arsenal of snap on tools to get anywhere including stubbies, knuckles, imperial, metric, 6 point and 12 point shorties and extended sizes etc. You have to take half the car apart to fix anything. They are designed to go 200K and then off to the scrap yard. Now they have complex touch screens. Really People! If driving with a cell phone is dangerous, what about farting with a touch screen to get your iTunes account to play the latest hits at 90 MPH in congested freeway traffic. Stupid! When they crash, they virtually fly apart in a crumpled mess. Even minor accidents are total losses! Now the insurance companies have to cover this joke so your insurance rates are thru the roof! The marketing folks put lip stick on this pig and get everyone to buy these plastic buggies on seven year contracts. The cars are done at the end of seven years so you get to buy another one. Then they now require you to use fluids with secret recipes and huge price tags. What happened to old fashioned lubricants and coolants? And now you can’t get a manual tranny! All computerized automatics! An automatic is a machinist’s wet dream! And they are at least twice or more expensive than an old gear box. Have we become so lazy that gear jamming is such an inconvenience? I am glad to be facing retirement as I don’t want any part of this nonsense!
@@devmeistersuperprecision4155 Um, I didn't subscribe to this newsletter, and modern cars are not that hard to work on. See... most any car repair channel on youtube.
Fascinating; great presentation of the problem.
Part of the issue here is the length of the beam and the Gauge of the steel. A long piece of thin steel in any configuration will be susceptible to torsion and buckle.
Absolutely, the length has a higher order power in the denominator, which makes the whole equation of torsional buckling basically governed by the length of the beam
7:25 - also any asymmetric loading over time, including while assembling, or dynamic loads.
I think especially assembly time loads and stresses that remain in the structure when fully assembled are easy to overlook.
id love to see this same test of the box tubing with a simple snug fit wooden center to renforce the wole beam
Thank you so much for the very helpful content. I appreciate your work! Also very understandable and real-life examples
Amazing content, i hope that our engineering proffesors teached us structural mechanincs like the way you do. Can i know what visualization program you used to modell the lateral tortional buckling of the steel beam in 4:58?
ABAQUS?
I love the small model you used. It makes novices like me understand the science. Thank you
The picture at 1:00 is worth a thousand words. When the beam is under load, both the top plates and bottom plates are *trying* to stretch, with the bottom wanting to stretch more. But, they _REALLY don't want to stretch!_
With the strap in the middle holding them together and keeping them from stretching different distances, there is only one way for them to go that keeps them from trying to stretch, so each can maintain as close to their original lengths as possible, which IS _the same length for each._
"Come on Baby, let's do the Twist!"
This video explains why Engineered Trusses in floors are WELL cross-braced, nowadays.
I have floor trusses there o.c.16" 2x4 constructed and 18" tall they span 28'. I have squash blocks around the rim with a 1.25 osb sub floor. So far so Good.
Great work! You continue to amaze me with your engineering skills. Keep up the good work!
Your loading is concentrated on very focused points along the beam. The load needs to be evenly distributed which helps the beam to function as it was intended.
Did he use the weakest looking beam he could find to make his point? Just look at that shitty ass beam, looks like you could break it by jumping on it.
@@lylemcdermott2566yes he did. He wanted to demonstrate failure safely. If he used a beam that would break at one ton it would take longer to load and failure could have caused damage to the person doing the loading. He got his point across and wasted less material with his demonstration.
@@lightningdemolition1964 but he drilled holes through the beam, didn't load it like it should have been loaded ( on the top) and beams aren't working that way in the real life scenario. For an engineer it's crap work.
It's a demonstration!!!
@@lylemcdermott2566 Not to mention he didn't use H beams. He bolted 2 U channels together, totally weakening it AND introducing weaknesses that wouldn't be there with a proper H beam. Did the same with the "box" beam, again 2 U channels bolted together to "create" a box beam.
Clear and concise. Brilliant article demonstrating great erudition
Thanks for all the citations. I think you definitely put a lot of time into making this video.
Thank you changw 🙏Reference [3], [4], and [5] are absolute gems. Luckily, as engineers, we have studied these books for many years which makes it a bit easier to provide citations but still, the animations and/or experiments do consume a lot of time.
A great explanation, good graphics, great information.
Neat, good job with the animations
I think the old RSJ beams were less prone to buckling and twisting because of the wedge design?.
Thanks for the demo info.
Used and wore out a wide flange beam In A wood spliter and was planning on doubling thickness of top flange where the spliter knife rides. Your thoughts please on this? Thankyou
I would suggest you add a plastic slider to the bottom of your splitter knife.
Excellent video, but the sound regarding the voice narration is terrible, have to crank up the volume then when an ad comes on it blasts so loud.
I think this is one reason we newer used open beam framework in the (north of) Nord for wooden buildings in historical times. Whan it was used in farmers buildings the beams are always square, never rectangular (in modern sense) in historical buildings here. Open beams where never used in living houses here, the roof on my old farmhouse is cladded with 2" plankings on both sides, and i think it is square beams under it in order to withstand the weight of wet snow.
Isn't the center the weak point? Closer to the end it would have more stay. This is good work, thanks.
For your beam tests, did you balance the weight distribution of the load, i.e. ensure that the load at each of the four points was equal?
As equal as possible, but of course, it can't be perfect
Great stuff for Saturday morning with a cup of coffee! Will drill deeper!
5:44
this is good one, thank you
That is why composite I-beams, and beams without fillets between the body and the flange are better. Because well filleted i-beams that are do not have wide flanges are far stronger across multiple axis.
Neither of those beams are typical of properly selected material thickness. These are very good examples of silly minimalism in materials selection.
The only thing one need do during design is to choose a failure mode that suits your needs. Because here is the thing, your test setup also failed to properly secure the ends of the beams you chose, so failure was inevitable because the beam never transmits tension to the points of support.
The problem seems much exacerbated by the excessive slenderness of the truss and the lack proper constrains at the extremities.
This allows the upper wing if the truss to become unstable, due to lack of axis tension, in a similar way as a free slender plank become unstable under vertical load faster then one which has working constrains.
What you are doing adding flanges to the side of the section is increasing the area of of it, and so reducing the slenderness of the truss itself.
I guess standard truss dimensioning, like the IPE system suffers less of the problem because it accounts for the issue.
I also imagine Baloon Framing is most vulnerable to this kind of problems
So a crystalline and straight beam would not experience this failure?
(Due to having no imperfections even down the the molecular level, assuming the molecular bonding angle and resulting crystal structure are not introducing a mechanical flaw)
Excellent presentation!
At 7:25 you say that the root cause is 'the imperfect manufacturing and construction process'. While that's part of it, I think it would be better described as the design being inherently unstable at high loadings. Vertical load on a deflected beam causes the beam to deflect further.
The initial deflection could come from manufacturing or installation, but even with a perfect beam installed perfectly, there are all kinds of things that could lead to deflection. Wind loading, earthquake loading, even just uneven roof loading (the next beam over has an AC unit sitting on it, this one doesn't, so it sags a few mm less and is picking up more load from one side) are all things that could push a beam off the perfectly balanced knife edge.
Absolutely agree, even a beam that's perfect down to the last molecule would experience this failure because any external actions (a flap of a butterfly's wings) would cause it to buckle.
Hello, the weight is 5.5 tons per meter. Which type of iron can bear this weight? Thank you
Seems like if two beams of equivalent material, one closed and one I beam, fail at such wildly different point loads when installed the same way; the use of open beams is only justifiable if there is a specific use for the flange or if the difference in manufacturing cost is really so great... but except at very large sizes it seems that both are priced pretty similarly.
The sections bend in the tranversal plane and generate horizontal loads, contrary to what is hypothesized in the calculations, hence the torsional buckling?
The calculations for the in-plane bending still hold. The only problem is that the torsional buckling happens first i.e. for a lower load (in slender sections) therefore the bending capacity does not govern. The torsional buckling happens due to the fact that the top of the beam is being compressed. If you imagine the compressed half of the beam as its own column then it's easy to see that it will want to "kick-out" of place. Opposite to that, the stretched part of the beam wants to stay where it is causing a twisting motion in the beam. I wouldn't say the calculations are wrong it's just a different failure mechanism.
I really appreciate these videos!! I'm designing concrete molds for my company and this has helped me choose the best pieces for each portion!
You did a fine job of explaining, but perhaps an explanation of the mathematics for the mathematically challenged. I would love to try and understand math more in-depth.
The math shown is nothing more than empirical values and equations, meaning they are simple formula that is contrived to match experimental test data, it's not intuitive at all and trust me, seeing it would not lead to any significant insight. I'll post a link below of a video that shows how beam deflection is derived based on simple geometry which shows what I think you're asking to see here. That sort of elegant mathematical derivation is not how the equations shown in this video work. ruclips.net/video/g8luMxP5Yo4/видео.html
Is top bar compressing bottom bar expanding, The bar in compression decides to return to its normal length and deflect sideways,
Interesting. I wonder if these equations could be made generic, if they aren't already, and have no bias for a particular metal so that the cross-section of a metal subject would be described and performed upon, analyzed and apply to all metals, if not all materials and substances, so there were no flaws in the equations and the same equation could be used for all cross-section shapes and materials. Thanks for the video.
Very interesting, thank you!
Even though they are not used in construction, I’d like to see how a cross shaped (like a X but verticale and horizontal instead of diagonal) cross section would hold up compared to the I beam shape or the box shape.
Intuitively I think having the wider width at the neutral point between tension and compression would add strength, but maybe not.
It would be substantially weaker than an equal weight box beam. A beam's resistance to bending is modeled using, among other properties, the 2nd moment of area of the beam's cross section. The result is that areas of a beam further from the neutral axis add more to the strength of the beam than area close to the neutral axis. This is why I-beams are so common: they're an extremely efficient shape for a strong but light beam.
@@markjacobson4248 but wouldn’t the extra width at the neutral point add more resistance to the twisting as the I beam is loaded?
If the load is placed centered on that neutral point it would be pulling the horizontal part down, along its weak axis, but also along the vertical part’s strong axis. As demonstrated, that load would cause the vertical part to deflect sideways, but the horizontal part would counter that by having a greater width at the neutral point.
I’m not an engineer, or even a student of engineering, so please don’t think my question of comment is stupid. I’m genuinely interested in this and am learning from it.
@@Linusgump As a simplified rule, the material in a beam that has the greatest contribution to the beam's resistance to deformation is the material that would experience the greatest local deformation to accomodate the overall deformation. Any material on the neutral axis of any deformation, whether twisting or bending, contributes nothing to strength because it doesn't deform at all. It's definitely not as straightforward with a complex motion like that in this video, where you have twisting around one axis and bending along two others, but the general rule still holds. A box beam would do a far better job of resisting deflection per weight. In general, the further material is from the neutral axis of any deformation, the more it matters. I-beams and H-beams are common design to maximize the material away from the neutral axis in the top and bottom flange, while the web between the flanges actually has minimal contribution to the bending strength of the beam. A hollow tube is a common design to maximize twisting resistance of a beam or structural member, as that maximizes the material away from the neutral axis of twisting, the central axis of the member. A box beam is a good compromise between both those designs, and is good for situations like this.
Even though I'm not an engineer, I could see right away that attaching stuff to an I-beam like in the thumbnail is a really bad idea. Even without calculations, the vertical part of the I-beam is clearly too tall and thin for that, and the horizontal parts are too thin as well, and so the slightest load can lacause deformations, which will eventually escalate to much serious buckling and failing. Either one redesigns the attachments or use different shaped (square sectioned) beams.
You see these on the highways every day. Flatbed step decks and heavy haul lowboy trailers. It's a ladder beam style construction with a crown in the center to combat bowing. This is why the run a spread axle and have bridge laws.
3:20 Now, THAT’s stupid scary. The smallest slip will cause the beam to slide out and chase someone. The yellow grips should be placed considerably wider. You can see it even start to wobble because the fulcrum is so small.
I am an EE but know now why structural engineers are brought in to do the heavy calcs for stuff like this now - for CPD it's brilliant as I never liked mech eng much - thanks 😁
I hope I have such a good teacher in my civil eng course
What program are you using for structural analysis?
Would this cause drywall to crack?
The torsional stiffness of an open beam is 100s of times smaller than a closed rectangular cross section beam of same dimensions, material and thickness. This means that stresses on open beams must be strictly longitudinal. Any other direction favors the closed beam, because it might be 25% heavier, but it is 100s times stronger.
That's exactly right! Also the message that the video was trying to convey.
I would also point out one little mistake, that the most stressed part of the I-beam at the time stamp ruclips.net/video/XSl7ZntK94E/видео.html would be on the other side of the web, because the compressive stresses from torsion and bending about both axes will sum right at that point. Besides that, very good educational video :) keep up the good work.
The critical most part of the beam is at the root of Junction of Web- flange in compression i.e. (tf+ Root radii) from the extreme end of the compression zone.
6:58 hi, must the member that restrain the compression side of I beam be diagonal member? Can a horizontal member ties to the compression face of I beam restrains it from moving laterally?
Horizontal braces are also fine. You can see this often in the roofs of Walmart, Costco, Superstore, etc. There are parallel trusses supported by larger girders. Usually, at the columns the truss is extended to the lower flange of the girder to peovide lateral stability. This is because there are large negative moments there, causing the girder's bottom flange to kick out.
The term is not difficult, but a beginner engineer must know how to become the best engineer
Had to unload a 50 foot I beam from a box trailer, 2 foot web. It took the better part of the morning to do it without twisting it and ruining the beam.
50 ft is fairly long; the length is the most important parameter, making the beam very suceptible to bending and torsion. Good job on your project 👏
the audio is really quite
i had to turn my headphones to 100%
What is the software he is using? thanks!
You should follow this with a video on a box truss section.
Isn't the instability of the beam better explained by the center of gravity of the load being offset to any degree from the vertical element of the beam? A box section achieves better stability by having a wider distance between vertical members, not from having fewer imperfections.
Sure, but that's only partly the reason. The susceptibility to lateral torsional buckling (LTB) is also due to the poor torsional resistance of the I-beam. Closed sections on the other hand are much better at resisting torsion and therefore have higher LTB resistance (not necessarily because of the center of gravity). A box beam with a slit along one side (open box section) would geometrically be very similar to a closed box beam but would not perform as well in LTB due to the fact of not being closed resulting in weak torsional resistance. The weak moment capacity in the lateral direction of the I-beam is a reason more why it is weaker than the box beam. Manufacturing imperfections are important but the essence of the weakness basically comes down to "it's just the way physics works". In retrospective, maybe in the video we made it sound like the manufacturing is the weak link when in fact it's the laws of universe 😅
You just need to tortionally stabilise the beam, whether through lateral bracing, or securing it to the floor plate above.
Excellent content
When the beams are connected, they won't buckle. The connection with the other beams, supports it.
Usually the people designing a structure will know if there will be any loading that would cause a beam to fail, you should already know this. They would design in crossbeams to ensure the beam wouldn't flex, I know this and I am a retired electrician.
Needs lateral bracing at compression top chord.
Engineers do understand the need to build superstatic structures, but they failed to recognize the need for a superstatic design for those beams.
With regards to lateral torsion, to say that the root cause is material imperfection is a red herring. Material had better have been inspected prior to installation, beams had better be tested before they are used in production. Especially per findings on this disaster where a beam was not inspected or tested:
en.wikipedia.org/wiki/Hyatt_Regency_walkway_collapse
Material is always imperfect, that is its nature, account for it, assume it to exist. Just like in electrical engineering, where the tiniest semiconductor imperfection in a balanced but unstable driven circuit causes oscillation, often as intended. If you intend it to be rigid, make it stiff in all directions. Never engineer in 2 dimensions because in 3-4 dimensions it will fail.
~One Electrical Engineer
Hi not amouse, this imperfection is so microscopic that no inspection can detect it and more importantly even if it could, it cannot be eliminated. Instead it is just assumed that it is there. It is the nature of our world, nothing is perfect. The problem is that this microscopic deflection gets amplified when an unbraced beam is loaded and results in lateral buckling. If in a parallel universe we could produce a perfect beam down to the Planck length it would still buckle laterally because the load is not perfectly centered and will always prefer one side over another. If somehow the load was also perfectly centered then the beam won't buckle laterally but it would take the flap of a butterfly's wings to initiate this catastrophic process. So in essence lateral torsional buckling does come down to imperfections and natural preference of the material to one side over the other. Engineer's prevent this by totally avoiding it. No design ever allows for lateral torsional buckling to occur in normal operating conditions.
In regards to your link, the Hyatt Regency walkway collapse is a very famous structural failure and we have studied and discussed it numerous times during our education and practice. A more rigorous inspection process and better communication most likely would have caught the error since it is absurdly obvious. But this is very different from the microscopic imperfections discussed in this video.
It's not really an imperfection; its simply the point that bends first.
This channel has potential, and please don't use airpods for recording audio
HAHAHAHA He really likes the word 'INTUITIVELY' take a drink everytime he says it and you are guaranteed to get hammered out of your head.
It is a great word 😅 I actually just searched the script and found only 3 matches for words starting with "intu." I'm not sure how many mentions are in the video, but in general, the voiceover follows the script. If it takes 3 drinks to get you hammered out of your head, then you can make a drinking game out of basically anything 😅
So, based on this analysis, why the popularity of I beams instead of box beams for construction?
@christianlibertarian5488 they are extremely strong when they can be braced or when the top of the beam is connected to a supporting slab, which prevents it from moving.
Nice video, but your volume is too low, I can barely hear you even with volume at maximum
Awesome video
Thanks !
How come these are used widely then?
As a fabricator I just make what the engineers say
But now I’m need to know why use i beam at all if a box beam for the same weight is stronger and harder to work with and more expensive
Hi Aussie, a box beam is only stronger for unrastrained applications. For most applications , W-sections can be laterally braced to prevent lateral-torsional buckling. In that case (restrained), the w-beam (or I-beam) is significantly more efficient at carrying bending loads.
At 6:32 you apply assimetrical load, isn't it?
There is probably some intrinsic asymmetry but nothing applied on purpose and hopefully only minimal.
@@TheEngineeringHub yes, you are absolutely right.
I designed a hybrid I beam that I think is an improvement .
Thanks
Closed channel beam ok, as long as you don't load it like the hyatt regency ones in 1981.
Your comparison test was NOT equal. The I beam has just the single thickness on the horizontal sections with ZERO fasteners strengthening them. While the box section has the horizontal sections doubled up with many fasteners strengthening it. It was not clear how the fasteners on the vertical sections compared. If they were equal, it made the box beam just that much stronger by the fact it had more material. The fact that the material is added where it strengthens the beam the most contribute to an unequal, and thus unfair test.
Spot welds would add no material, but would be just about impossible to do on the box beam. Regardless, it would still do more to strengthen the box beam than it does the I beam.
A fairer test would be I beams and box beams of the same height, width and the same weight. Made that way, not assembled from C sections.
Honeycomb beams reign supreme for load bearing in construction designs. You can't change my mind!
You sound like a student, give it a couple years and you realize there's no such thing as "best", it's all relative.
So in short, the I-beam was used in the wrong configuration and with the wrong load...?
Please be continue
That formula again.. 😱😱🤣🤣
Brings back memories of 9/11 and the root cause of the failing top floors.