At the time the video was shot, it was no longer a crack, but a complete separation of the two sections. The effects of the motion (every time the cars rolled on that section of tracks) played a major role on the stress of the structure (dynamic or live loading; fatigue). There was a picture taken the previous week by someone else that actually showed the crack near the connection of the tubes. In about a week or so, this led to a complete fracture. That picture was shown on national news.
.... sure looks like a fatigue failure..... it broke! .....and does not appear to have stretched first...... well, could also have an inclusion in the steel but, surely someone tested for that!
You know I have to wonder about the quality of the materials being used to build these things. I'm not saying that's definitely the case, but if it's happening to more than one ride over the course of time, is it possible they could be getting substandard materials at the fabricating plant, and it might not show up in their testing?
Same issue happened on Thunderbird at Holiday World last year. It's more common than people think, and it's really not a huge issue. It closes the ride for a while, but its still safe
i am a ironworker out of clevland and worked on and built most of these rides at cedar point. The milleneum force sheared solid billet tubing on the downhill and broke plenty of horizontal bracing in the tall towers on the big hill. The raptor is the same engineering pipe columns the same track connections.1 bolt you could not get in on the uphill the chain drive was in the way. Their was already enough bolts 1 inch metric i think Intamin desiged raptor. The vertigo one of the columns broke above a connection, they siad wind shear caused it. I could go on and on i worked their from 1991- 2010 it got to cold in the winter i wqent to florida.
@@coastaku1954ha, y’all’s intimidator is awesome and definitely beats our intimidator but fury is one of the best coasters in the world, I mean intimidator 305 is awesome and all but it’s not that special.
@@aldenweaver8237 Was Fury 325 known for causing people to black out? was Fury 325 modified because of how strong the first turn was? Fury literally has dead spots and not much airtime, All meh floater, weak turns and a SHIT TON of Rattle, literally gives me a headache. I305 is my Number 1 and Fury is my Number 8, behind The Beast, Orion and Steel Vengeance, just to name a few
I certainly hope they review all the other rides at the park now. Its amazing that this was noticed by an ordinary park-goer instead of the maintenance staff, hopefully it broke only recently and hasn't been like that for a week or so!
Safety critical structure is typically required to be damage tolerant. That means that the design engineers had to consider load redistribution due to a catastrophic failure such as this and ensure that the remaining load path can handle the loads (temporarily) until the defect could be discovered and repaired. If it's anything like air vehicle structures.. all the adjacent, redundant structure probably needs to be repaired or replaced as well due to the overload condition that the failed support caused.
@@christosvoskresye correct... Hence why I said it can only sustain damage tolerance (DT) conditions "temporarily" and that adjacent structure likely needs to be repaired or replaced as well. Part of the point of DT is so that localized failures like this can be spotted and fixed before they become deadly. It's about giving more time between inspections before catastrophic failures occur. The overload condition that results on adjacent structure indeed accelerates it's own failure rate and the DT load path is broken as the failed support is probably the intended DT load path for those adjacent members. Thus failure of a 2nd adjacent support would likely result in catastrophic failure of the track. FYI when engineers design custom repairs for in service structures, they can redesign the load path and fix design flaws with the original design. So a "repair" doesn't mean bringing the structure back to the same configuration as when the ride was first built.
Ryan The Ride Mechanic, a coaster tech was saying he suspects that the footer on the diagonal piece may have sunk a bit, putting additional stress on that support. What do you think about that?
Well, the only other possibility is the ride was overstressing the part and that's almost impossible as the B&M's as well as most other roller coasters, are overbuilt from an engineering standpoint.
Very good possibility. I'm not a civil engineer, but anytime the restraints change, the loads will change. A sinking of the support would indeed change the restraints and loads.
In my opinion, the brace footer is unlikely to sink more than the vertical footer (straight down). The brace is in tension under design loads, thus uplift on that footer. I agree that the weld failure is most-likely the result of weld quality. Nearly impossible to actually inspect a weld penetration without an x-ray, and I doubt they x-ray every weld during fabrication.
@@MentoredEngineer Viaducts with multiple columns are also considering differential settlements in design. The stresses arising in overdetermined structures are quite big. However differential settlements are usually not really an issue with foundations so close to each other (although cannot fully be excluded).
Add weld would increase the heat effective zone causing the material to weaken. While welding these joints there should of been pre & post heat temperatures checked as well as during. If you get the material to hot the material is as good as cardboard
I believe this was a heat treatment problem. People are saying welding problems. I disagree. Probably either the piece was made of pure trash, or not properly treated.
@MrSolidBlakeGaming correct. There had to be a wps for the job with the correct wire, gas, voltage etc pre heat and post heat as well as inner pass temperature. There's a lot of weld in that part.
We are talking about low strength, structural carbon steels. Something in the realms of a S275 or A36, with wall thicknesses of about maybe 15mm. As a civil and welding engineer, I can assure you one thing: it's close to impossible to weaken mild carbon steels, even with bad welding parameters. Only thing how you can really hurt such steels would be rapid cooling with water directly after welding. We are not talking about high wall thicknesses or high strength structural steels, which get their strength by quenching & tempering or by thermomechanical rolling processes. And what some of the commenters probably don't know: bad welds mainly affect the static strength, but for fatigue loads the simple existence of welds is already sufficient to lower the strength of a part. Assuming a S275 steel, the material can bear 275N/mm² before starting to yield. The strength of the welds themselves is usually even higher. The allowable stress range in fatigue for this connection is (highly depending on the geometry) is roughly somewhere in between 30 to 80 N/mm²
They announced yesterday that a new support is currently being fabricated and will be delivered next week. I am doubting, however, that it will have any engineering improvements. I’m gonna guess they’ll get the ride up and running ASAP with replacement parts of the current design and will be inspecting it frequently while they investigate what caused the failure and engineer a long term solution, then during the off season we’ll see the ride partially disassembled and the new solution actually implemented for next season.
@jaanfo3874 I know they announced it yesterday but I said it 5 days ago. I'll also bet they do add a gusset for extra strength or change the angles that the joining pipes are connected to. B&m don't like issues and have no problem making corrections that's why they're typically highly reliable
Civil and welding engineer here, I have to disagree with quite a few of the points here. I have done research in hollow section joints. The stress concentration usually on the weld toes (weld cap) for most geometries is much higher than the stress concentration at the weld root. So for most of the joint geometries, it simply does not matter at all, if you carry out a full penetration weld or a fillet weld, as the cracking will start on the outside on one of the weld toes. You can even predict which one of the toes is most likely to start cracking, look at CIDECT Design Guide 8. Two other things to consider: not sure about AISC 360 and AWS D1.1, but other codes a) require a full penetration weld at least at this crown point of the connection and b) in case of fillet welds require to make them full strength, i.e. the fillet weld must have the same strength as the base material. Due the varying stiffness and therefore stresses around the perimeter of the weld, you cannot tell without FEA how high the stresses at any point are, hence go full strength everywhere. And avoiding hollow section joints? Good luck. Those are standard joints with more than enough research behind them to also create formulas for assessing their strength without having to use FEA. And if designed correctly, those joints will not fail under static loading. Or, to be more precise: fail with a probability less than 1:1,000,000. BTW: you concentrated solely on the design. You completely neglected the two other parts of the whole equation: a) manufacturing. A weld flaw in that area and the whole desing goes boom. And don't think NDT would help, because if there's no requirement, NDT is done on statistical basis, aka a certain percentage. And depending on the method of NDT and type of flaw, your detection rate is ranging from zero to about in best case (phased array ultrasonic testing) about 80%. In case manual UT was done, you can expect detecion rates of about 40-60%. b) most engineers do not know, that especially for fatigue design, you can't design with 100% probability of survival. The Wöhler lines for fatigue design are based on a statistical evaluation using a normal distribution with 97.5%/97.7% probability of survival. I once did a series of tests with approx 50-60 test specimen. And guess what? I had one outlier which was below the 97.5% survival line. So this would also be an option why this thing has started cracking: pure coincidence. And that's why especially for fatigue loaded parts, regular inspections are mandatory.
The way the broken column barely moves with the heavy car going past it at speed seems to indicate that this was already a failed structure from day one and the light to moderate stress over time finally revealed the flaw, fortunately it wasnt 100% load bearing or it could have possibly failed catastrophically within weeks of it first being built.
Glad you looked into this. What aboutnatural frequency causing the crack. My experience with vibrating conveyors, I came across many failures where the vibrating frequency, which was at or about the natural frequency of the metal structure caused the failure. At one point one of the major users required the entire piece of equipment to under go analysis prior to approving drawings. I think it will be found to be more than just a bad weld or defective metal. time will tell, one way or another.
I doubt it. A substantial amount of a conveyor's load is from operating around the natural frequency. With Fury, the cart running is 99% of the load and you are looking at a fairly low fatigue count compared to a conveyor which may have several load cycles a second.
The best part is the engineer who designed the coaster to begin with designed it in a way that would result in catastrophic prevention if something that this would happen. Funny how that worked out
They could install two vertical concrete filled rebar reinforced columns behind the existing column and brace them together with a collar and attach the track to the collar to take the stress off the original column, they could do the same with the remaining columns that see high stress and prolong the life of the ride for decades.
When I saw this my first thought was that the angular support should be on the opposite side whereby pushing towards the force of the train vs pulling away from the support. My layman's eye visualized that force being transferred down the length of the angular tube toward the ground whereby dispensing the force over a greater area.
@@MentoredEngineer I've ridden the Fury 325 and many other Giga coasters. I don't think safety and peoples lives care about aesthetics and parking. I know mine surely does not. This was caused by either bad planning, poor design or wishful thinking on the part of Cedar Fair and Bolliger & Mabillard. From my perspective a repair on the existing design is a Band-Aid on a broken leg. They need to redesign the support. If that intrudes on their parking and they deem that intolerable then they need to reroute the track to provide either less gravitational stress or better support IMO. They also need to look at the portions of the ride that have similar stresses and supports and consider altering them as well. There's light poles in the parking lot, hang a light on it lol! Can't be any worse than that Taurus that caught fire and took another 11 cars with it.🤦♂
And it took them all of 2 weeks to get B&M's factory in Batavia to get another support made. People forget that the blueprints are still readily available and another one can be fabricated quite quickly. The thing to worry about is... what caused this in the first place on a coaster that is only 8 years old. Did they have a design flaw that took 8 years to manifest itself or has something about the ground The coaster sits on changed somehow. It's easily a 5 ton vehicle empty. Add 32 people of varying weights, shapes, and sizes, and that load quickly jumps to 8 tons or more. Their engineering department better find out what the real cause of this catastrophic structural integrity failure was, because I have yet to hear it from someone that actually knows for a fact, not just engineering speculation.
Dropping the plate down to that 45* tube seems like the most cost effective option. I’ve heard that back in the day when sex was safe and racing was dangerous, tube chassis cars (particularly of exotic metals) would have the tubes pressurized with nitrogen and a gauge was installed somewhere in the chassis. Idea being if the pressure reads very low or zero, you have a crack. Would it be possible to do that with these tubes? That away, at a glance, maintenance could see if everything is ok or needs closer inspection.
@@MentoredEngineer Not common at all, only time I saw or better used air pressure was in in a research project to determine Wöhler curves for thick walled hollow section joints.
It seems weird to me that the diagonal support wasn't on the other side, bracing the column in opposition to the sideways load of the train going through that turn. Do you think that would have made a difference, having the the diagonal support on the outside of the turn vs the inside?
I am not an engineer but a coaster enthusiast who tries to understand things...I actually agree and was thinking the same thing. The loads at that portion I am sure are large given the speed and weight of the trains. The track and support braces only support the loads for a split second until the train has cleared...It likely cracked due to tension forces...not compression...
It appears like the vertical column failed but as was mentioned in the video the weld joint between the vertical and diagonal column created a very weak joint which after repeated stresses over the years eventually caused a fracture which steadily grew over time.
If the support is outboard, the ride takes up more space. And generally, steel is stronger in tension than compression. But I agree, at the failure location it would be a bearing point, but you could still see the stress in the top side of the connection.
@@zachheitman6247 For the steel itself, it does not matter if in compression or tension, the yield strength is the same for both. However ANY structural part in compression also faces buckling problems, which reduce the strength of the whole element. Does not matter which material, steel, concrete, wood, aluminium or anything else. However steel is more prone to such stability problems due to it's high strength and therefore slender construction compared to low strenght materials.
And to add fuel to the 🔥....... *The Fury 325 roller coaster, which can take 32 passengers at a time, was shut down on July 1 by the Carowinds amusement park after staff were notified of a crack at the top of one of the ride's steel pillars. And according to North Carolina Labor Commissioner Josh Dobson, the crack may have been visible as early as 10 days before the ride's closure. "It looks like maybe six to 10 days prior, some pictures had been taken that shows the beginning of the crack, and then by obviously last Friday, the thing was completely severed," he told the Associated Press on Friday. - courtesy of Insider*
The problem is the techs do not do regular track inspections. I work at a large park and track inspections are conducted everyday. Some areas daily some weekly depending on the stress put on that section. These rides will unfortunately always end up eventually having a Crack and needs to be repaired they key is catching it early.
@MentoredEngineer @MentoredEngineer how is that contradicting? I said regular track inspections some areas on the track are daily if they are known to be under a lot of stress and the areas that are not under high stress is weekly. For example every night you will see a tech repel slowly down one section of a ride because that particular section has a daily PM that has to be done or the ride can not open for the day. On another day you will see a tech repeling down that same section while other techs are in a couple of high reaches inspecting other sections of the track. Because thoes are on weekly PM's. Every ride is different and has pm's that are at different intervals based on that particular ride. In some cases you will have a portion of the track that has daily, weekly, and monthly pm's. Thoes PM's are designed to go over thoes areas in detail. During thoes checks if an issue is found that is when QC is called to inspect the area with special tools and if needed call weld out to excavate and weld the section that is in question. Regardless of the daily, weekly, and monthly PM's which are specific to certain areas on the track and is a detailed inspection the entire track is looked at by a couple of people everyday but this is a general visual inspection.
Exactly what you said about inspecting the beams before and after, I would ad replace all the hardware that holds the track that have been compromised too.
Looks like what coasters need is a robot that can drive itself along the track, with cameras and other sensors to check the track, and robot arms with cameras that reach out, around, behind etc to examine the supports from connection to the track down to a few feet away. Movement and distortion could be checked for in a simple way. Weld target markers at various locations. When the robot gets to specific locations along the track, and the robot arms move to specific locations, if the target markers are not *exactly* centered in the cross-hairs (calibrated at first robot run after completion of the coaster build) then something that shouldn't move has moved. With all the target markers precisely located, a computer could have a 3D model of the coaster, feed it the marker position difference measurements from the robot and see how the movement has affected the coaster. As with FEA software, the distortion could be exaggerated as a visual aid to see it better. To ensure that the robot itself hasn't become out of spec, it should have a very overbuilt test fixture that *cannot warp*. The fixture should have various test markers to which the robot rolls and moves its arm(s). If things don't align, it's the robot out of whack. Service the robot, recalibrate to the test fixture, then check the coaster. The test fixture could be in the garage where the robot us kept between its scheduled inspection runs. To ensure robot body to support alignment, there could be a pin or ring or other attachment the robot could clamp onto so it can't slip on the rails. Like with multi-axis industrial robots "slotted into" assembly line positions and programmed to use the same tools humans formerly wielded, a coaster inspection robot wouldn't miss looking at everything it's supposed to look at, needs no pay or health insurance, never skips a scheduled inspection because it had a fight with its girlfriend, never will miss a crack because of a bad hangover headache...
You didn’t come up with anything new. That’s already a thing. They already have robots that run along the track and inspect the rollercoaster. They even have inspection cars that humans ride in. I’d imagine they could be upgraded though.
I suspect that the design was fine for static loads. They underestimated the effect of cyclic loading (think of bending a paper clip until it breaks). Fatigue strength of mild steel drops quickly as the number of cycles increases. I don’t know how old this ride is, but at 1 car per minute, 12 hours per day, 250-300 days per year, it wouldn’t take too many years to get to millions of cycles. It’s even greater when you account for the number of axles per car as each axle passes over a point it’s separate event as far as the structure is concerned. Watch the track, the next time you are stopped at a railway crossing. Also, it would be very easy for a slight misalignment during fabrication. A couple of millimeters or 1/10’s of a degree would make a significant difference.
Things happen over time. Likely stress fracture. Repeated side load. They will have to check this thing with an X-ray annually if they decide to just replace it like it was. They may actually want to scan ALL the beams now to see if this is happening internally on any of the others.
There's always a better way. Having ridden this ride at least 100 times, I hate to see this happen, just grateful no one was injured , Carowinds is one of the safest amusement parks in America, and im sure will offer an excellent solution to remedy/ repair this flagship thrill ride!
Good anlysis. I was thinking stress concentration failure due to poor weld design/job, possibly resulting in the weld fatiguing above the endurance limit for the material.
There are other tasks that take place on a weekly, monthly, and annual basis. At least once per year, most parks X-ray their track or use magnetic scanners to check for metal stress or welds that need attention. Also, all vehicles are completely disassembled, inspected, and rebuilt at least once a year.Jan 31, 2011 - Google Soooo.......
My comment wasn't directed towards you. But since you asked; what I posted was how amusement park rollercoasters should be maintained. It was obvious to me that coaster couldn't have been inspected recently. Definitely not that day before the park opened. The following supports my thought. "In the video, the roller coaster zooms by, moving the cracked beam and widening the crack’s gap back and forth. Jeremy Wagner, who was at the Carowinds Amusement Park in Charlotte with his kids, recorded the video before reporting it to park staff. The park’s newest and tallest ride was eventually shut down later on in the day. Wagner first showed the film to an elderly parking attendant because he was already outside the park. The man said he didn’t see anything when the roller coaster cars came back around and revealed the crack. But Wagner insisted that the park needed to shut down the ride immediately. He said it took the fourth person he encountered after he got into guest services to do anything about the situation. “There was no sense of urgency,” Wagner said. “That was what really concerned me.” - courtesy of Concho Valley and WJZY I read the article early this morning.
*The Fury 325 roller coaster, which can take 32 passengers at a time, was shut down on July 1 by the Carowinds amusement park after staff were notified of a crack at the top of one of the ride's steel pillars. And according to North Carolina Labor Commissioner Josh Dobson, the crack may have been visible as early as 10 days before the ride's closure. "It looks like maybe six to 10 days prior, some pictures had been taken that shows the beginning of the crack, and then by obviously last Friday, the thing was completely severed," he told the Associated Press on Friday. - courtesy of Insider*
Well, all of that would be true if this were static loading.... which it clearly is not! Steel does not crack or break unless it has a fatigue failure due to the number of cycles and load seen. Normal steel testing shows that, under a pure static tensile load, it stretches before it breaks. Also, the support is cantilevered-loaded so, the front weld has a tremendous load, as you pointed out, and being load-cycled about every 30 seconds..... and unless someone tested that part, we don't even know the stress or strain on that front weld.... And, since momentum = mass x velocity², you have a pretty good idea of the force applied. But, with no other data... this looks like a fatigue failure......
If you look closely you can see that the vertical tube was notched into then had the angle pipe welded into it. If the notch had not been cut into the vertical pipe i would not have cracked the way it did. The weakest point is always the weld if you look closely the weld is fine. Its the vertical pipe that cracked meaning the vertical pipe was taking on an unstable load to begin with. Its bad design the vertical weld should have snapped not the pipe cracked horizontally.
In my design of weldments course (mentoredengineer.com/product/design-of-weldments-class/), we teach that the normal stress in the weld should be less than 60% of the weld wire tensile stress. So yes, the weld is always the weak part
Did you happen to see the photo that a women posted, taken the Saturday prior? It shows the crack had formed (pardon my lay terms) on the "slanty", angled part, but hadn't yet spread to the vertical part yet.
Not that big of a deal to inspect these days. You have excellent drones with 4K footage that can assist with these inspections. Those drones can be used to continuously film EVERY joint and weld without someone having to climb the ride every single day.
Brandon with Theme Park Predictions says they have many drones doing inspection as well as men walking the track. However, I'm sure there are many areas like this one where it will be difficult to inspect unless you are within arms length.
Why would you need a drone when this was visible from the parking lot several days before the fracture and shut-down of the ride? Blaming this on a "hard to inspect" area is nonsense. A visual inspection from the ground would certainly have caught this - since a visual inspection from the parking lot DID!
You can tell by the way the outside face of the column 'tore' that that was the last part of the failure, which supports your theory that the failure emanated from the weld.
What I meant is that the trains tend to move a little faster in the afternoon once the wheels and bearings have heated up a little. The physics of it provide for a max speed of 97-98 mph
Possibly, although maintenance of these systems might be worse than regular inspection. I'm sure there would be a lot of false positives due to malfunction.
I'm curious - How safe is the structure with the break? Obviously you'd not run the coaster as soon as it's noticed, but I'm curious how much margin of safety there would be for the coaster losing an entire support. I'd assume it would be able to run safely and that no guests were ever at risk, but I don't know what kind of safety margins are built in. I'd also assume that since it's putting much higher loads on the supports next to it, those would have a much shorter life before they became unsafe, but would that potentially be on the order of days or years?
The "factor of safety" or margin of safety depends on the design and what the manufacturer built into it. Where I work, we build electric motors and depending on which component it is or what is being used we use a safety factor of anything from 2 up to over 10. generally when it comes to safety and lifting, we are around 5-10, but other components we may be at 2 because they may not be subject to much stress beyond that. FYI, a safety factor of 2 means it can handle twice the stress it was designed for. So if i wanted to design something to handle say, 1000 psi, with a safety factor of 2, I would design it to handle 2000 psi. In regards to something like this failure, there could be a multitude of issue. Design is one of them, but also the manufacturing of the part, the material, the weld etc. could have been bad from the beginning. My opinion here is not that the design was necessarily bad, so much as it could have been a material issue or a weld / manufacturer issue. But who knows, engineers make mistakes sometimes.
@tjsengle great insight and explanation. I hope the design factor used is higher than 2:1. The ANSI 360-10 standard has the endurance limit of the joint at 7,000 psi and 4500 psi for a Category D and E weld, which would probably limit it further than a normal design factor.
As mentioned, I would be very concerned what was going on at the adjacent supports since if the connection was inadequate at the failed support, now the adjacent supports are taking additional loading and I wonder how adequate are those connections. As the owner, I would be weary of the entire design and would want an independent review of the design with a specific focus on the same typical detail of supports that failed. This is a serious failure for a structure that is not very old which to me indicates under design rather than steel fatigue induced failure.
Ryan the ride mechanic said that the crack wasn’t much of a safety hazard and it takes multiple footers to crack to cause safety problems, it probably just put more stress in the track possibly making that track piece age a lil faster but that’s it.
Great question. Yes and No. They could do that, but it kind of defeats the point. The springs would absorb energy from the system and reduce the velocity as it goes by. The stiffer a coaster is, the faster it would be. I had to change the wheels to be stiffer on my backyard coaster so that it would get all the way around the track.
Would not be helpful. Springs could dampen the effects of local, small impacts (bumps etc.), but not the overall nominal loads coming from the train due to outwards acceleration. Those forces are still there and cannot be reduced, except for lowering the speed. What also needs to be taken into account: if this support is less stiff, the loads coming from the train will be transferred to the neighbouring columns, putting higher loads on those.
Thank you for your expert explanation. I have a question. Is it possible that this crack can lead to finding many more mistakes (maybe x-raying the structure with a portable x-ray machine mounted on a tall crane for high areas) that the coaster would have to be torn down and rebuilt from scratch?
I doubt that the coaster will need to be torn down. B&M has a long record of excellent structures. There will be a fully inspection of the structures using all sorts of nondestructive testing (maybe even x-ray). There is a growing theory that the diagonal support foundation may have sunk causing excess loads on this joint.
Is there a reason there is no beam angled outward. I don’t know the exact speed or radius of that turn, but wouldn’t you want at least some horizontal bracing to oppose that force?
It failed. So the safety margin wasn’t a factor of 2-3. Which to me indicates many or most joints may be at margins of 1-1.5. And public perception and insurance company risk assessment would IMO require a serious rework and X-ray inspection of EVERY weld gusset fastener. Recertification is gonna be hard. As it should be.
If that type of joint and weld is so bad, why does B&M use it so much? So many of the supports on all B&M coasters are similar with a vertical tube and an angled tube attached near the top. Also, what sort of damage would be required for there to be a complete structural failure? My guess and hopes would be that multiple adjacent supports would need to fail (i.e. any two supports could fail like this as long as there is a "good" support between them).
Evidence would support the joint is not "so bad". B&M has been using this type of joint for 30 years. While we see that the weld is the weak part, it could be that the foundation shifted adding extra loads
Unfortunately it was not made clear in this video that Corey does not actually know how the joint design is. It is a pure assumption that a fillet weld was used. And even if a fillet weld was used, hollow section joints usually have higher stress concentrations on the weld toes rather than the weld root, i.e. it would not make any difference if you use a fillet or a full/partial penetration weld. Assuming that the welds themselves have sufficient strength.
Could the ground cause be overload? I mean, when the coaster was constructed, they might have counted with a median load of say 90 kg per person, meaning they calculated all support structures with wagons good for say, 1T per 10person wagon. When people generally becomes heavier due to eating too much McD, lets say median weight of 120 kg per person, there will be overweight of 200 kg, which translated to lateral dynamic forces is about 800kg-1T. I checked a table of G-forces and Fury have 4G, thats a overweight of 800 kg on that support pillar. No wonder it is subject to metal fatigue, as it was not built for those loads. Maybe amusement parks should start weighting their passengers to make sure that the wagons both are evenly loaded, but also so no wagon becomes overweight? Or impose weight limits, a scale that will show a green for OK weight, or red for too low/high weight for safe travel. Or even, at entry, you weight, and you will then get a coded bracelet that has a encrypted QR-code telling which rides you can take.
Static design according to ASTM 2291 requires to assume a 300lb load for a single person in static load cases. For fatigue load cases 170lb are assumed, simply due to the assumption that not all trains will be loaded with overweight persons in each and every run during the ride's lifetime. How often does this happen during the lifetime, a train fully loaded with 300lb people? Maybe never, maybe a few times, that's nothing in the grand scheme.
@@johnscaramis2515 But can we be sure that they have adhered to ASTM 2291 standard? and 170 lb is a bit low even for fatigue, since the median weight is higher today. People are more fatty and eat more McD, they become obese, and even if you take a good mix of light and heavy people, you are still going to blast the 170 lb limit. They should IMHO start weighting all passengers and add to a tally. When next person weighting would blast the max total weight for a train, they should send off that train even with a few empty chairs, and have next train to be filled. this means every train might go with like 2-4 empty chairs because too many obese persons boarded the train, but then max weight may not be exceeded. Sending off trains with empty chairs also means obese persons don't get last in line all the time just because you are "trying to fill each train", instead the rule becomes fair, if train is full, its full weight-wise even if theres empty chairs, and we send it off. Could be arranged with a scale plate just where you stand before boarding, and computer adds each passenger's weigt to a total tally, if the person currently standing on scale, would cause train to go overweight, train is sent off, then he is first to board next train instead.
@@sebastiannielsen I am working in a company that checks ride calculations. And I can assure you, companies like B&M will always follow those requirements, otherwise they won't get their necessary paperwork. It's also a matter of liability. I assume that the people in the working groups of ASTM are no idiots. The codes are reworked every few years, if there is a significant change in weight, they will incorporate it. The last revision of the code is from 2021 or 2022. What most people and also many engineers don't know: for static design it's either fail or withstand. For fatigue, it's a summation of individual small damages per each run. You may have a slightly higher stresses during one run, but then you also have empty runs (e.g. daily empty run before opening), which cause close to no damage, but are counted as runs. Unfortunately many engineers are simply working with it as it is similar to static design. And if we look on the resistance side, things get even more complicated. If you take two parts and test them with the same load, one might withstand 100,000 cycles, whereas the other can withstand 1,000,000 cycles. Or maybe 2,000,000. There are so many other uncertainties in fatigue design, which are not predictable, passenger loads are only one part of those. That's why there are regular inspections, to cover for those uncertainties. So even with your plan, a fatigue crack might happen. Even if you would limit the ride only to 170lb persons, a crack may happen. And who knows, in the end it might have been simply a flaw in the weld.
I feel that might be unnecessary. Worst case is they fix this support and add another on each side. My guess is the footers sank a little causing extra load on the joint.
Everyone calling this a “design flaw” how about you see the support was fully detached and the ride was still running flawlessly. This rides an engineering masterpiece quit over reacting.
Sad thing thing when it was brought to the attention of park officials, they wouldn't shut it down, It took calling the fire dept and they shut it down.
I've never seen angled supports angled in the same direction as the forces. I'm no engineer but it looks like bad design and made this crack inevitable. They should be angled against the forces.
The post is designed to have some flex, your solution is to remove the flex. That will ultimately result in a catastrophic failure, where in this case, the other supports acted a redundant support, and took the load. Two things will happen. The entire support will be replaced By B&M, possibly more as they took unusual stress for an unknown time period or the ride will be removed. Hopefully the Park President and Maint. Director will be fired for a culture of lax inspections. This was likely the result in three things, not two. Weather as metal expands in the heat, unanticipated stresses over time and what will be eventually stated, an unforeseen failure in the mfg process of the support. The latter being easier to state than design flaw as there are other coaster made similarly by B&M across the country. Cedar Fair will downplay the risks and safety and hope the final story is buried. The Park President and Maint Director will be left untouched as no one was hurt and the park did not get sued.
Cracks on coasters happen more often than you think. Look at the Bat or Big Bad Wolf. Your opinion on the flex is also incorrect. Military aircraft, like cargo planes, have wings that move 10-12 between on the ground and in the air. Commercial airliners maybe have 1-2 ft of sway under normal operating loads. We don't see airliners falling out of the sky because they are stiff. It is all in your design goals. There are probably way more than 3 things responsible for the failure and the investigation will find them all including human factors (operations, management and inspections) as well as the mechanics. The temperatures in question here are not in a range that will effect the stresses in any significant way.
@@MentoredEngineer Please read what I wrote and not what you think I wrote. The post needs flexibility, otherwise it would snap like a toothpick. Your solution appears to remove flex and movement. A stiffer system is unhealthy. Watch every roller coaster, they bend and move. I also attempted imply the weather was a factor, not the total answer. 'Cracks appear more often than you think,' yes, aware but they are usually caught. This one was not. Also, this coaster is a baby at 8 years old. I can tell you stories about what appeared to be catastrophic failures which went unnoticed, and no one got hurt. Right now every Park is re-inspecting everything, including Six Flags, all Cedar Fair, Merlin, Universal and hopefully Disney to name a few. However you are still safer on ANY roller coaster than in the car ride to and from the park. Thanks for the videos.
Seems like they should have a support that’s “pushing” rather than “pulling”, right? For example, the support looks like this: /|, would it not be better as this: |\ ?
As we are talking about fatigue induced failure, putting the support on the other side would have one advantage: the weld toe would be in compression instead of in tension. While the design codes for welded parts only asses the stress range and neglect the direction of stresses and the mean stress (many good reasons to do so), doing fatigue tests on my own, I know that welds under compression perform slightly better than welds under cyclic tensile stresses. However the whole argument goes down the drain, if weld flaws are involved.
It is doubtful that the fundamental design of the support is the issue. Many are saying that the diagonal tube's footer sunk a little which would add extra load to the joint. B&M has a great track record with this design.
First off, B&M knows what they are doing and the round tube joint has a proven track record. However, I prefer square tubing because it mates together better. This also allows for the easy addition of doublers and gussets that can help ease the transition. Square tubing does cost more though.
@@MentoredEngineer However the fatigue strength of square hollow section joints is much lower than circular hollow sections, hence you would need thicker cross-sections to compensate.
A laser scanner probably wouldn't have found anything except maybe the settling of a foundation if that exists and contributed to the failure. The only way to find this earlier is pretty much a manlift to do mag particle testing. That being said this crack didn't go to a full break in one day. They definitely need to improve their visual inspection process to make sure all joints and connections are subject to a close visual inspection (using a zoom lens or drone) probably on at least a weekly basis.
@@danbert8 You can get some pretty good results on integrity with a modern scanner - just not early enough to stop the loss of integrity from happening. I agree - the scanner becomes helpful when its already "too late", but they could have probably caught this months ago and have already mitigated the issue. I just assumed that parks would be scanning footers and such regularly and comparing the model data to make sure there arent issues. I wish I was still working for Trimble - I'd be having a field day with this.
@@MentoredEngineer I'm guessing they're referring to using lasers for very precise measurements to create a 3D model of position. That would, of course, be non-destructive, and any (unexpected/unallowable) change in position would require further investigation to determine why something isn't where it should be.
I love when the general public trys to argue with me, the amount of people who thinks they just gunna drop a new support in or weld it back up in just a few weeks. these people have no idea about construction. inspections alone take weeks if not months. and we dont know yet weather they will have to retro fit new supports or even foundations on other parts of the ride. I just hope they do this right
Indeed. The investigation will take 2-3 months minimum. B&M and Carowinds will be working on repair while that is going on and then implement it quickly. It may be open for Scarowinds later.
Would not change anything, because the problem is a the point where the column head meets the weld of the diagonal. Maybe a better explanation: if you mount an antenna on the roof of the building and the anchor points between antenna and roof fail, stengthening the building will not help.
The way that support is designed, there is shear force working at that weld joint. Would it not be better to have the support brace on the outside so that the joint is now in compression when it is stressed?
Read where the amusement park said they inspect daily (like all amusement parks say).....calling B.S The crack is to large/ long to be just 2023 season problem......
Maybe. It takes a long time for a crack to start, but once it is going it can fail very quickly. Structures aren't a daily inspection because it would be too time consuming. Woodies only have to have the track walked each day, not the entire structure.
You lost me at “inadequate design”. You jumped to that conclusion before even considering manufacturing errors, bad welding, inadequate NDT, ground settling, or other non-design-related possible issues. Several engineers with many years of experience spent a lot of time designing this ride, and it had to have been signed off by PE’s in order to get built in the first place. Saying the entire support needs to be redesigned is a HUGE leap. I’m not saying it’s impossible that there was a design error, but all things considered it’s very unlikely. I know this incident is fresh and is surely getting you clicks, but please consider the possible effects of what you’re saying here. As an engineer you should understand the ethical implications of spreading public mistrust.
Apparently you missed my disclaimer. The load has not changed since the day it opened and it broke; therefore, the design was inadequate. The investigation that will ensue will identify the root causes. Rest assured, we will learn more from this one broken support than all the rest of the successful supports.
Even if that would be allowed, if we are talking about fatigue cracks, those develop over time, hence won't be there when the ride is assembled. But if you think of checking each column each and every single day before opening, you need staff for that. Or time. Let's see, the track is about 6600 ft long, spans between columns are roughly every 20ft, probably less. Makes 330 columns. How much staff do you want to put on this task? And would you be willing to pay for this staff?
@mentoredengineer can you comment on the next likely occurrence if the crack had not been found? I know this is just prediction, but my husband and I are big coaster enthusiasts, we just left Hershey this morning! We have a small build in our backyard ❤ Weve discussed, what could have happened if it were continuously running as is. I have no idea, but he seems to think it could have ran a lot longer without any real catastrophic event.
Well, the load was distributed to the nearest supports on each side so there was no immediate danger. However, this increased loading will decrease the lifespan of those supports. Kudos on the backyard coaster. I'm making a backyard coaster database. Feel free to enter yours. mentoredengineer.com/backyard-roller-coaster-database/
Well, if the track were to break or even just kink, the train would derail or come to a very abrupt stop. Most likely this would cause injury and/or death.
@@MentoredEngineerAt least it's nice to know that beam isn't the only support holding up the track, or the only support holding up the train for that matter.
Your analysis ignores a very obvious issue with this design: Expansion in the track and no joints to accommodate it against a rigid support system. Better welds will not eliminate that stress and ultimate mode of failure.
Im a structural welder. I seen the close up photos. Lots of under cut in the top edge of weld. And thats where it broke. That welder shouldn't be on that job.
LOL - Cedar Fair has some real short sighted people high up. I dealt with a similar situation where 1980s fiber optics were banned and we had to wait for everyone that was around at the time to retire before reintroducing fiber-optics
This is not the only one that cracks were found at. Most of this steel made today is garbage. Any ride you see closed is a good sign that cracks were found at stress points.
Except that cracking usually does not happen within the base material itself, but at the weld toe or inside the weld. Having no clue, but having a strong opinion and spreading it. A typical youtube commenter.
At the time the video was shot, it was no longer a crack, but a complete separation of the two sections. The effects of the motion (every time the cars rolled on that section of tracks) played a major role on the stress of the structure (dynamic or live loading; fatigue). There was a picture taken the previous week by someone else that actually showed the crack near the connection of the tubes. In about a week or so, this led to a complete fracture. That picture was shown on national news.
.... sure looks like a fatigue failure..... it broke! .....and does not appear to have stretched first...... well, could also have an inclusion in the steel but, surely someone tested for that!
You know I have to wonder about the quality of the materials being used to build these things.
I'm not saying that's definitely the case, but if it's happening to more than one ride over the course of time, is it possible they could be getting substandard materials at the fabricating plant, and it might not show up in their testing?
Considering I've never seen a failure like this on any other B&M, I wouldn't rule out a mistake from the contractor.
Monster at Walygator Parc had the same issue years ago
Same issue happened on Thunderbird at Holiday World last year. It's more common than people think, and it's really not a huge issue. It closes the ride for a while, but its still safe
Silver Bullet at Knott's Berry Farm failed like this at the track spine.
A mistake would be getting back on that thing. It's not sound and the design is extremely bad
i am a ironworker out of clevland and worked on and built most of these rides at cedar point. The milleneum force sheared solid billet tubing on the downhill and broke plenty of horizontal bracing in the tall towers on the big hill. The raptor is the same engineering pipe columns the same track connections.1 bolt you could not get in on the uphill the chain drive was in the way. Their was already enough bolts 1 inch metric i think Intamin desiged raptor. The vertigo one of the columns broke above a connection, they siad wind shear caused it. I could go on and on i worked their from 1991- 2010 it got to cold in the winter i wqent to florida.
This is an amazing breakdown thank you so much for this
Glad it was helpful!
SO SAD! Driving my friend across the country later this month and was going to surprise him with a vist to CW just for Fury. Nice video!
While it sucks, at least it’s being addressed so people don’t get hurt.
Take him to Kings Dominion, Intimidator 305 blows Fury 325 WAY out of the water
Carowinds isn’t a terrible park without fury, we still have after burn, copperhead strike and intimidator.
@@coastaku1954ha, y’all’s intimidator is awesome and definitely beats our intimidator but fury is one of the best coasters in the world, I mean intimidator 305 is awesome and all but it’s not that special.
@@aldenweaver8237 Was Fury 325 known for causing people to black out? was Fury 325 modified because of how strong the first turn was? Fury literally has dead spots and not much airtime, All meh floater, weak turns and a SHIT TON of Rattle, literally gives me a headache.
I305 is my Number 1 and Fury is my Number 8, behind The Beast, Orion and Steel Vengeance, just to name a few
I certainly hope they review all the other rides at the park now. Its amazing that this was noticed by an ordinary park-goer instead of the maintenance staff, hopefully it broke only recently and hasn't been like that for a week or so!
Carowinds has a history for hiding things until the news drags it out in public
It's literally insane that it was able to get that far
Check this out. th-cam.com/video/W40VNnH1dSQ/w-d-xo.html This coaster had so many issues in the 3 seasons it was open.
Safety critical structure is typically required to be damage tolerant. That means that the design engineers had to consider load redistribution due to a catastrophic failure such as this and ensure that the remaining load path can handle the loads (temporarily) until the defect could be discovered and repaired.
If it's anything like air vehicle structures.. all the adjacent, redundant structure probably needs to be repaired or replaced as well due to the overload condition that the failed support caused.
@@christosvoskresye correct... Hence why I said it can only sustain damage tolerance (DT) conditions "temporarily" and that adjacent structure likely needs to be repaired or replaced as well.
Part of the point of DT is so that localized failures like this can be spotted and fixed before they become deadly. It's about giving more time between inspections before catastrophic failures occur.
The overload condition that results on adjacent structure indeed accelerates it's own failure rate and the DT load path is broken as the failed support is probably the intended DT load path for those adjacent members. Thus failure of a 2nd adjacent support would likely result in catastrophic failure of the track.
FYI when engineers design custom repairs for in service structures, they can redesign the load path and fix design flaws with the original design. So a "repair" doesn't mean bringing the structure back to the same configuration as when the ride was first built.
Ryan The Ride Mechanic, a coaster tech was saying he suspects that the footer on the diagonal piece may have sunk a bit, putting additional stress on that support. What do you think about that?
Well, the only other possibility is the ride was overstressing the part and that's almost impossible as the B&M's as well as most other roller coasters, are overbuilt from an engineering standpoint.
^^
Very good possibility. I'm not a civil engineer, but anytime the restraints change, the loads will change. A sinking of the support would indeed change the restraints and loads.
In my opinion, the brace footer is unlikely to sink more than the vertical footer (straight down). The brace is in tension under design loads, thus uplift on that footer. I agree that the weld failure is most-likely the result of weld quality. Nearly impossible to actually inspect a weld penetration without an x-ray, and I doubt they x-ray every weld during fabrication.
@@MentoredEngineer Viaducts with multiple columns are also considering differential settlements in design. The stresses arising in overdetermined structures are quite big.
However differential settlements are usually not really an issue with foundations so close to each other (although cannot fully be excluded).
Add weld would increase the heat effective zone causing the material to weaken. While welding these joints there should of been pre & post heat temperatures checked as well as during. If you get the material to hot the material is as good as cardboard
I believe this was a heat treatment problem. People are saying welding problems. I disagree. Probably either the piece was made of pure trash, or not properly treated.
@MrSolidBlakeGaming correct. There had to be a wps for the job with the correct wire, gas, voltage etc pre heat and post heat as well as inner pass temperature. There's a lot of weld in that part.
@@MrSolidBlakeGaming Structural parts are usually not heat treated, except you want to machine them. And columns are definitely not heat treated.
We are talking about low strength, structural carbon steels. Something in the realms of a S275 or A36, with wall thicknesses of about maybe 15mm.
As a civil and welding engineer, I can assure you one thing: it's close to impossible to weaken mild carbon steels, even with bad welding parameters.
Only thing how you can really hurt such steels would be rapid cooling with water directly after welding.
We are not talking about high wall thicknesses or high strength structural steels, which get their strength by quenching & tempering or by thermomechanical rolling processes.
And what some of the commenters probably don't know: bad welds mainly affect the static strength, but for fatigue loads the simple existence of welds is already sufficient to lower the strength of a part.
Assuming a S275 steel, the material can bear 275N/mm² before starting to yield. The strength of the welds themselves is usually even higher. The allowable stress range in fatigue for this connection is (highly depending on the geometry) is roughly somewhere in between 30 to 80 N/mm²
Very good and sad video but that line "Hey I'm wearing the same shirt" is a daily saying of all engineers lol
So true
At least it didn't have a pocket protector. Dangerous on a coaster.
😂😂😂 true!
There is a photo from 24 Jun, by a visitor, showing the partial crack.
SEEN FROM THE GROUND.
A drone vid inspection every week would be cheap enough.
So glad someone noticed this crack in the beam.
The track is fabricated at the clemont steel works in Cincinnati. Might not take as long as you'd think for a piece to be made
Let's hope so!
They announced yesterday that a new support is currently being fabricated and will be delivered next week. I am doubting, however, that it will have any engineering improvements. I’m gonna guess they’ll get the ride up and running ASAP with replacement parts of the current design and will be inspecting it frequently while they investigate what caused the failure and engineer a long term solution, then during the off season we’ll see the ride partially disassembled and the new solution actually implemented for next season.
@jaanfo3874 I know they announced it yesterday but I said it 5 days ago. I'll also bet they do add a gusset for extra strength or change the angles that the joining pipes are connected to. B&m don't like issues and have no problem making corrections that's why they're typically highly reliable
Civil and welding engineer here, I have to disagree with quite a few of the points here.
I have done research in hollow section joints. The stress concentration usually on the weld toes (weld cap) for most geometries is much higher than the stress concentration at the weld root. So for most of the joint geometries, it simply does not matter at all, if you carry out a full penetration weld or a fillet weld, as the cracking will start on the outside on one of the weld toes.
You can even predict which one of the toes is most likely to start cracking, look at CIDECT Design Guide 8.
Two other things to consider: not sure about AISC 360 and AWS D1.1, but other codes a) require a full penetration weld at least at this crown point of the connection and b) in case of fillet welds require to make them full strength, i.e. the fillet weld must have the same strength as the base material. Due the varying stiffness and therefore stresses around the perimeter of the weld, you cannot tell without FEA how high the stresses at any point are, hence go full strength everywhere.
And avoiding hollow section joints? Good luck. Those are standard joints with more than enough research behind them to also create formulas for assessing their strength without having to use FEA. And if designed correctly, those joints will not fail under static loading. Or, to be more precise: fail with a probability less than 1:1,000,000.
BTW: you concentrated solely on the design. You completely neglected the two other parts of the whole equation:
a) manufacturing. A weld flaw in that area and the whole desing goes boom. And don't think NDT would help, because if there's no requirement, NDT is done on statistical basis, aka a certain percentage. And depending on the method of NDT and type of flaw, your detection rate is ranging from zero to about in best case (phased array ultrasonic testing) about 80%. In case manual UT was done, you can expect detecion rates of about 40-60%.
b) most engineers do not know, that especially for fatigue design, you can't design with 100% probability of survival. The Wöhler lines for fatigue design are based on a statistical evaluation using a normal distribution with 97.5%/97.7% probability of survival. I once did a series of tests with approx 50-60 test specimen. And guess what? I had one outlier which was below the 97.5% survival line. So this would also be an option why this thing has started cracking: pure coincidence.
And that's why especially for fatigue loaded parts, regular inspections are mandatory.
ok great!
The way the broken column barely moves with the heavy car going past it at speed seems to indicate that this was already a failed structure from day one and the light to moderate stress over time finally revealed the flaw, fortunately it wasnt 100% load bearing or it could have possibly failed catastrophically within weeks of it first being built.
Glad you looked into this. What aboutnatural frequency causing the crack. My experience with vibrating conveyors, I came across many failures where the vibrating frequency, which was at or about the natural frequency of the metal structure caused the failure. At one point one of the major users required the entire piece of equipment to under go analysis prior to approving drawings. I think it will be found to be more than just a bad weld or defective metal. time will tell, one way or another.
I doubt it. A substantial amount of a conveyor's load is from operating around the natural frequency. With Fury, the cart running is 99% of the load and you are looking at a fairly low fatigue count compared to a conveyor which may have several load cycles a second.
The best part is the engineer who designed the coaster to begin with designed it in a way that would result in catastrophic prevention if something that this would happen. Funny how that worked out
Yes, they did a great job! A support completely broke, but the ride stayed intact and no one was hurt
You deserve the views you're getting on this video. Great channel, I've been using some info for the last month and a half on my log splitter build.
Much appreciated! Thanks. The log splitter what what inspired me to start the channel in the first place.
They could install two vertical concrete filled rebar reinforced columns behind the existing column and brace them together with a collar and attach the track to the collar to take the stress off the original column, they could do the same with the remaining columns that see high stress and prolong the life of the ride for decades.
A lady took a picture of the beginning signs of the crack about a week prior, on june 24th
When I saw this my first thought was that the angular support should be on the opposite side whereby pushing towards the force of the train vs pulling away from the support. My layman's eye visualized that force being transferred down the length of the angular tube toward the ground whereby dispensing the force over a greater area.
It would make more sense to have that angled away so the joint would be in compression. However, that would put it in the parking lot.
@@MentoredEngineer I've ridden the Fury 325 and many other Giga coasters. I don't think safety and peoples lives care about aesthetics and parking. I know mine surely does not. This was caused by either bad planning, poor design or wishful thinking on the part of Cedar Fair and Bolliger & Mabillard.
From my perspective a repair on the existing design is a Band-Aid on a broken leg. They need to redesign the support. If that intrudes on their parking and they deem that intolerable then they need to reroute the track to provide either less gravitational stress or better support IMO. They also need to look at the portions of the ride that have similar stresses and supports and consider altering them as well.
There's light poles in the parking lot, hang a light on it lol! Can't be any worse than that Taurus that caught fire and took another 11 cars with it.🤦♂
I haven't heard so much talk about crack since I got out of rehab.
And it took them all of 2 weeks to get B&M's factory in Batavia to get another support made.
People forget that the blueprints are still readily available and another one can be fabricated quite quickly.
The thing to worry about is... what caused this in the first place on a coaster that is only 8 years old.
Did they have a design flaw that took 8 years to manifest itself or has something about the ground The coaster sits on changed somehow.
It's easily a 5 ton vehicle empty. Add 32 people of varying weights, shapes, and sizes, and that load quickly jumps to 8 tons or more.
Their engineering department better find out what the real cause of this catastrophic structural integrity failure was, because I have yet to hear it from someone that actually knows for a fact, not just engineering speculation.
Dropping the plate down to that 45* tube seems like the most cost effective option. I’ve heard that back in the day when sex was safe and racing was dangerous, tube chassis cars (particularly of exotic metals) would have the tubes pressurized with nitrogen and a gauge was installed somewhere in the chassis. Idea being if the pressure reads very low or zero, you have a crack. Would it be possible to do that with these tubes? That away, at a glance, maintenance could see if everything is ok or needs closer inspection.
That is a great idea. I have no idea if that is a common practice.
Sex has always been dangerous.
@@MentoredEngineer Not common at all, only time I saw or better used air pressure was in in a research project to determine Wöhler curves for thick walled hollow section joints.
Me who saw the person filming while i was on the ride*
You really rode it with the broken support? Could you tell?
Thank god 🙏🏽 🙏🏽🙏🏽
Why you lying 💀 what’s the point
I rode it 3 times yesterday. No you couldn’t tell anything was wrong with it.
Do you remember if there was a pot hole at that section of the ride?
It seems weird to me that the diagonal support wasn't on the other side, bracing the column in opposition to the sideways load of the train going through that turn. Do you think that would have made a difference, having the the diagonal support on the outside of the turn vs the inside?
I am not an engineer but a coaster enthusiast who tries to understand things...I actually agree and was thinking the same thing. The loads at that portion I am sure are large given the speed and weight of the trains. The track and support braces only support the loads for a split second until the train has cleared...It likely cracked due to tension forces...not compression...
It appears like the vertical column failed but as was mentioned in the video the weld joint between the vertical and diagonal column created a very weak joint which after repeated stresses over the years eventually caused a fracture which steadily grew over time.
If the support is outboard, the ride takes up more space. And generally, steel is stronger in tension than compression. But I agree, at the failure location it would be a bearing point, but you could still see the stress in the top side of the connection.
@@zachheitman6247 For the steel itself, it does not matter if in compression or tension, the yield strength is the same for both. However ANY structural part in compression also faces buckling problems, which reduce the strength of the whole element. Does not matter which material, steel, concrete, wood, aluminium or anything else. However steel is more prone to such stability problems due to it's high strength and therefore slender construction compared to low strenght materials.
Precisely.
And to add fuel to the 🔥.......
*The Fury 325 roller coaster, which can take 32 passengers at a time, was shut down on July 1 by the Carowinds amusement park after staff were notified of a crack at the top of one of the ride's steel pillars. And according to North Carolina Labor Commissioner Josh Dobson, the crack may have been visible as early as 10 days before the ride's closure. "It looks like maybe six to 10 days prior, some pictures had been taken that shows the beginning of the crack, and then by obviously last Friday, the thing was completely severed," he told the Associated Press on Friday. - courtesy of Insider*
Thanks AI
The problem is the techs do not do regular track inspections. I work at a large park and track inspections are conducted everyday. Some areas daily some weekly depending on the stress put on that section. These rides will unfortunately always end up eventually having a Crack and needs to be repaired they key is catching it early.
eh? So you contradicted yourself. Do the techs do regular daily and weekly inspections are not?
@MentoredEngineer @MentoredEngineer how is that contradicting? I said regular track inspections some areas on the track are daily if they are known to be under a lot of stress and the areas that are not under high stress is weekly. For example every night you will see a tech repel slowly down one section of a ride because that particular section has a daily PM that has to be done or the ride can not open for the day. On another day you will see a tech repeling down that same section while other techs are in a couple of high reaches inspecting other sections of the track. Because thoes are on weekly PM's. Every ride is different and has pm's that are at different intervals based on that particular ride. In some cases you will have a portion of the track that has daily, weekly, and monthly pm's. Thoes PM's are designed to go over thoes areas in detail. During thoes checks if an issue is found that is when QC is called to inspect the area with special tools and if needed call weld out to excavate and weld the section that is in question. Regardless of the daily, weekly, and monthly PM's which are specific to certain areas on the track and is a detailed inspection the entire track is looked at by a couple of people everyday but this is a general visual inspection.
Footers need to be examined to make sure they have not started sinking, putting extra stress on braces.
True
Exactly what you said about inspecting the beams before and after, I would ad replace all the hardware that holds the track that have been compromised too.
Good point. In the industry I worked in we would do that on critical joints after an issue like this.
Looks like what coasters need is a robot that can drive itself along the track, with cameras and other sensors to check the track, and robot arms with cameras that reach out, around, behind etc to examine the supports from connection to the track down to a few feet away. Movement and distortion could be checked for in a simple way. Weld target markers at various locations. When the robot gets to specific locations along the track, and the robot arms move to specific locations, if the target markers are not *exactly* centered in the cross-hairs (calibrated at first robot run after completion of the coaster build) then something that shouldn't move has moved. With all the target markers precisely located, a computer could have a 3D model of the coaster, feed it the marker position difference measurements from the robot and see how the movement has affected the coaster. As with FEA software, the distortion could be exaggerated as a visual aid to see it better.
To ensure that the robot itself hasn't become out of spec, it should have a very overbuilt test fixture that *cannot warp*. The fixture should have various test markers to which the robot rolls and moves its arm(s). If things don't align, it's the robot out of whack. Service the robot, recalibrate to the test fixture, then check the coaster. The test fixture could be in the garage where the robot us kept between its scheduled inspection runs. To ensure robot body to support alignment, there could be a pin or ring or other attachment the robot could clamp onto so it can't slip on the rails.
Like with multi-axis industrial robots "slotted into" assembly line positions and programmed to use the same tools humans formerly wielded, a coaster inspection robot wouldn't miss looking at everything it's supposed to look at, needs no pay or health insurance, never skips a scheduled inspection because it had a fight with its girlfriend, never will miss a crack because of a bad hangover headache...
Not a bad idea. Can I have it?
@@MentoredEngineer go for it
You didn’t come up with anything new. That’s already a thing. They already have robots that run along the track and inspect the rollercoaster. They even have inspection cars that humans ride in. I’d imagine they could be upgraded though.
I suspect that the design was fine for static loads. They underestimated the effect of cyclic loading (think of bending a paper clip until it breaks). Fatigue strength of mild steel drops quickly as the number of cycles increases. I don’t know how old this ride is, but at 1 car per minute, 12 hours per day, 250-300 days per year, it wouldn’t take too many years to get to millions of cycles. It’s even greater when you account for the number of axles per car as each axle passes over a point it’s separate event as far as the structure is concerned. Watch the track, the next time you are stopped at a railway crossing.
Also, it would be very easy for a slight misalignment during fabrication. A couple of millimeters or 1/10’s of a degree would make a significant difference.
Yeah, they definitely haven't built roller coasters before and obviously didn't know that there would be repetitive load/unload events.
Things happen over time. Likely stress fracture. Repeated side load. They will have to check this thing with an X-ray annually if they decide to just replace it like it was. They may actually want to scan ALL the beams now to see if this is happening internally on any of the others.
Good point
There's always a better way. Having ridden this ride at least 100 times, I hate to see this happen, just grateful no one was injured , Carowinds is one of the safest amusement parks in America, and im sure will offer an excellent solution to remedy/ repair this flagship thrill ride!
This will be back up and running before we know it.
would def like to know what the inspection cycle of this area is and when the last DNI was.
Why is the 45° supporting-tube in front, and not in the back to take more of the stress from the weld?
Good anlysis. I was thinking stress concentration failure due to poor weld design/job, possibly resulting in the weld fatiguing above the endurance limit for the material.
Great point!
There are other tasks that take place on a weekly, monthly, and annual basis. At least once per year, most parks X-ray their track or use magnetic scanners to check for metal stress or welds that need attention. Also, all vehicles are completely disassembled, inspected, and rebuilt at least once a year.Jan 31, 2011 - Google
Soooo.......
So... what? I missed your point.
My comment wasn't directed towards you. But since you asked; what I posted was how amusement park rollercoasters should be maintained. It was obvious to me that coaster couldn't have been inspected recently. Definitely not that day before the park opened. The following supports my thought.
"In the video, the roller coaster zooms by, moving the cracked beam and widening the crack’s gap back and forth. Jeremy Wagner, who was at the Carowinds Amusement Park in Charlotte with his kids, recorded the video before reporting it to park staff. The park’s newest and tallest ride was eventually shut down later on in the day. Wagner first showed the film to an elderly parking attendant because he was already outside the park. The man said he didn’t see anything when the roller coaster cars came back around and revealed the crack. But Wagner insisted that the park needed to shut down the ride immediately. He said it took the fourth person he encountered after he got into guest services to do anything about the situation. “There was no sense of urgency,” Wagner said. “That was what really concerned me.” - courtesy of Concho Valley and WJZY
I read the article early this morning.
*The Fury 325 roller coaster, which can take 32 passengers at a time, was shut down on July 1 by the Carowinds amusement park after staff were notified of a crack at the top of one of the ride's steel pillars. And according to North Carolina Labor Commissioner Josh Dobson, the crack may have been visible as early as 10 days before the ride's closure. "It looks like maybe six to 10 days prior, some pictures had been taken that shows the beginning of the crack, and then by obviously last Friday, the thing was completely severed," he told the Associated Press on Friday. - courtesy of Insider*
Very entertaining and fun to watch. Thanks for posting this.
Glad you enjoyed it!
Well, all of that would be true if this were static loading.... which it clearly is not! Steel does not crack or break unless it has a fatigue failure due to the number of cycles and load seen. Normal steel testing shows that, under a pure static tensile load, it stretches before it breaks. Also, the support is cantilevered-loaded so, the front weld has a tremendous load, as you pointed out, and being load-cycled about every 30 seconds..... and unless someone tested that part, we don't even know the stress or strain on that front weld.... And, since momentum = mass x velocity², you have a pretty good idea of the force applied. But, with no other data... this looks like a fatigue failure......
If you look closely you can see that the vertical tube was notched into then had the angle pipe welded into it. If the notch had not been cut into the vertical pipe i would not have cracked the way it did. The weakest point is always the weld if you look closely the weld is fine. Its the vertical pipe that cracked meaning the vertical pipe was taking on an unstable load to begin with. Its bad design the vertical weld should have snapped not the pipe cracked horizontally.
In my design of weldments course (mentoredengineer.com/product/design-of-weldments-class/), we teach that the normal stress in the weld should be less than 60% of the weld wire tensile stress. So yes, the weld is always the weak part
Did you happen to see the photo that a women posted, taken the Saturday prior? It shows the crack had formed (pardon my lay terms) on the "slanty", angled part, but hadn't yet spread to the vertical part yet.
Not that big of a deal to inspect these days. You have excellent drones with 4K footage that can assist with these inspections. Those drones can be used to continuously film EVERY joint and weld without someone having to climb the ride every single day.
Brandon with Theme Park Predictions says they have many drones doing inspection as well as men walking the track. However, I'm sure there are many areas like this one where it will be difficult to inspect unless you are within arms length.
Why would you need a drone when this was visible from the parking lot several days before the fracture and shut-down of the ride? Blaming this on a "hard to inspect" area is nonsense. A visual inspection from the ground would certainly have caught this - since a visual inspection from the parking lot DID!
You can tell by the way the outside face of the column 'tore' that that was the last part of the failure, which supports your theory that the failure emanated from the weld.
I’d say it’s a bad heat treatment. Not a welding problem
I doubt these supports were heat treated. It would add no value to the steel material being that it a standard piping which is a low alloy.
Once the crack starts, you will have less material to carry the load. At a certain point, the material will just tear off as you pointed out.
4:00 I feel u hit the nail on the head they did speed up the ride from 95 to I think 105-110
What I meant is that the trains tend to move a little faster in the afternoon once the wheels and bearings have heated up a little. The physics of it provide for a max speed of 97-98 mph
Wouldn't it be less prone to cracking if the angled post that extends under the ride was placed in the opposite direction away from the ride? 🤔
Yup, but lands in the middle of the parking lot.
Stress points could have sensors added to set off alarms when parts completely separate.
Possibly, although maintenance of these systems might be worse than regular inspection. I'm sure there would be a lot of false positives due to malfunction.
@@MentoredEngineer True, but it would also make someone get a close look at the joint.
I'd count that as a win.
I'm curious - How safe is the structure with the break?
Obviously you'd not run the coaster as soon as it's noticed, but I'm curious how much margin of safety there would be for the coaster losing an entire support.
I'd assume it would be able to run safely and that no guests were ever at risk, but I don't know what kind of safety margins are built in.
I'd also assume that since it's putting much higher loads on the supports next to it, those would have a much shorter life before they became unsafe, but would that potentially be on the order of days or years?
The "factor of safety" or margin of safety depends on the design and what the manufacturer built into it. Where I work, we build electric motors and depending on which component it is or what is being used we use a safety factor of anything from 2 up to over 10. generally when it comes to safety and lifting, we are around 5-10, but other components we may be at 2 because they may not be subject to much stress beyond that. FYI, a safety factor of 2 means it can handle twice the stress it was designed for. So if i wanted to design something to handle say, 1000 psi, with a safety factor of 2, I would design it to handle 2000 psi. In regards to something like this failure, there could be a multitude of issue. Design is one of them, but also the manufacturing of the part, the material, the weld etc. could have been bad from the beginning. My opinion here is not that the design was necessarily bad, so much as it could have been a material issue or a weld / manufacturer issue. But who knows, engineers make mistakes sometimes.
@tjsengle great insight and explanation. I hope the design factor used is higher than 2:1. The ANSI 360-10 standard has the endurance limit of the joint at 7,000 psi and 4500 psi for a Category D and E weld, which would probably limit it further than a normal design factor.
I would not ride it FD3 comes to mind
As mentioned, I would be very concerned what was going on at the adjacent supports since if the connection was inadequate at the failed support, now the adjacent supports are taking additional loading and I wonder how adequate are those connections.
As the owner, I would be weary of the entire design and would want an independent review of the design with a specific focus on the same typical detail of supports that failed. This is a serious failure for a structure that is not very old which to me indicates under design rather than steel fatigue induced failure.
Ryan the ride mechanic said that the crack wasn’t much of a safety hazard and it takes multiple footers to crack to cause safety problems, it probably just put more stress in the track possibly making that track piece age a lil faster but that’s it.
wow! tons of quick work on this. Thanks for the info
You bet!
Excellent presentation.👏👏
Thank you!
Great video, however you have a large amount of static in your audio, might want look into addressing it.
The top of the track seems to rest naturally a few cm to the right of the support. I wonder if the support needs to move to the right a bit as well?
Since the cars rolling by exert stress on the joint, could they install some kind of spring system at that location that would absorb the stress?
Great question. Yes and No. They could do that, but it kind of defeats the point. The springs would absorb energy from the system and reduce the velocity as it goes by. The stiffer a coaster is, the faster it would be.
I had to change the wheels to be stiffer on my backyard coaster so that it would get all the way around the track.
Would not be helpful. Springs could dampen the effects of local, small impacts (bumps etc.), but not the overall nominal loads coming from the train due to outwards acceleration. Those forces are still there and cannot be reduced, except for lowering the speed.
What also needs to be taken into account: if this support is less stiff, the loads coming from the train will be transferred to the neighbouring columns, putting higher loads on those.
Thank you for your expert explanation. I have a question. Is it possible that this crack can lead to finding many more mistakes (maybe x-raying the structure with a portable x-ray machine mounted on a tall crane for high areas) that the coaster would have to be torn down and rebuilt from scratch?
I doubt that the coaster will need to be torn down. B&M has a long record of excellent structures. There will be a fully inspection of the structures using all sorts of nondestructive testing (maybe even x-ray). There is a growing theory that the diagonal support foundation may have sunk causing excess loads on this joint.
Is there a reason there is no beam angled outward. I don’t know the exact speed or radius of that turn, but wouldn’t you want at least some horizontal bracing to oppose that force?
It would have put the support in the parking lot.
It failed. So the safety margin wasn’t a factor of 2-3. Which to me indicates many or most joints may be at margins of 1-1.5.
And public perception and insurance company risk assessment would IMO require a serious rework and X-ray inspection of EVERY weld gusset fastener.
Recertification is gonna be hard. As it should be.
If that type of joint and weld is so bad, why does B&M use it so much? So many of the supports on all B&M coasters are similar with a vertical tube and an angled tube attached near the top. Also, what sort of damage would be required for there to be a complete structural failure? My guess and hopes would be that multiple adjacent supports would need to fail (i.e. any two supports could fail like this as long as there is a "good" support between them).
Evidence would support the joint is not "so bad". B&M has been using this type of joint for 30 years. While we see that the weld is the weak part, it could be that the foundation shifted adding extra loads
Unfortunately it was not made clear in this video that Corey does not actually know how the joint design is. It is a pure assumption that a fillet weld was used.
And even if a fillet weld was used, hollow section joints usually have higher stress concentrations on the weld toes rather than the weld root, i.e. it would not make any difference if you use a fillet or a full/partial penetration weld. Assuming that the welds themselves have sufficient strength.
Weekly drone inspection should be a thing.
This is so sad but we just have to brace ourselves because it’s going to take time to get it fixed.
Lets not crack under pressure!
It could also be the 45 degree angle tube it’s foundation shifted or sank a little
In most cases don't they usually dig the holes for the footers down to bedrock so as to prevent settling of the footers after construction?
Could the ground cause be overload? I mean, when the coaster was constructed, they might have counted with a median load of say 90 kg per person, meaning they calculated all support structures with wagons good for say, 1T per 10person wagon. When people generally becomes heavier due to eating too much McD, lets say median weight of 120 kg per person, there will be overweight of 200 kg, which translated to lateral dynamic forces is about 800kg-1T. I checked a table of G-forces and Fury have 4G, thats a overweight of 800 kg on that support pillar. No wonder it is subject to metal fatigue, as it was not built for those loads.
Maybe amusement parks should start weighting their passengers to make sure that the wagons both are evenly loaded, but also so no wagon becomes overweight? Or impose weight limits, a scale that will show a green for OK weight, or red for too low/high weight for safe travel. Or even, at entry, you weight, and you will then get a coded bracelet that has a encrypted QR-code telling which rides you can take.
Static design according to ASTM 2291 requires to assume a 300lb load for a single person in static load cases. For fatigue load cases 170lb are assumed, simply due to the assumption that not all trains will be loaded with overweight persons in each and every run during the ride's lifetime.
How often does this happen during the lifetime, a train fully loaded with 300lb people? Maybe never, maybe a few times, that's nothing in the grand scheme.
@@johnscaramis2515 But can we be sure that they have adhered to ASTM 2291 standard? and 170 lb is a bit low even for fatigue, since the median weight is higher today. People are more fatty and eat more McD, they become obese, and even if you take a good mix of light and heavy people, you are still going to blast the 170 lb limit.
They should IMHO start weighting all passengers and add to a tally. When next person weighting would blast the max total weight for a train, they should send off that train even with a few empty chairs, and have next train to be filled.
this means every train might go with like 2-4 empty chairs because too many obese persons boarded the train, but then max weight may not be exceeded.
Sending off trains with empty chairs also means obese persons don't get last in line all the time just because you are "trying to fill each train", instead the rule becomes fair, if train is full, its full weight-wise even if theres empty chairs, and we send it off.
Could be arranged with a scale plate just where you stand before boarding, and computer adds each passenger's weigt to a total tally, if the person currently standing on scale, would cause train to go overweight, train is sent off, then he is first to board next train instead.
@@sebastiannielsen I am working in a company that checks ride calculations. And I can assure you, companies like B&M will always follow those requirements, otherwise they won't get their necessary paperwork. It's also a matter of liability.
I assume that the people in the working groups of ASTM are no idiots. The codes are reworked every few years, if there is a significant change in weight, they will incorporate it. The last revision of the code is from 2021 or 2022.
What most people and also many engineers don't know: for static design it's either fail or withstand. For fatigue, it's a summation of individual small damages per each run. You may have a slightly higher stresses during one run, but then you also have empty runs (e.g. daily empty run before opening), which cause close to no damage, but are counted as runs. Unfortunately many engineers are simply working with it as it is similar to static design.
And if we look on the resistance side, things get even more complicated. If you take two parts and test them with the same load, one might withstand 100,000 cycles, whereas the other can withstand 1,000,000 cycles. Or maybe 2,000,000.
There are so many other uncertainties in fatigue design, which are not predictable, passenger loads are only one part of those. That's why there are regular inspections, to cover for those uncertainties.
So even with your plan, a fatigue crack might happen. Even if you would limit the ride only to 170lb persons, a crack may happen.
And who knows, in the end it might have been simply a flaw in the weld.
20k views in one day. Congrats Corey! - Adam
Thanks Adam. My best video ever. Keep me up to date with the investigation.
Depending on where in the layout this break occurred they could possibly reprofile the ride layout to eliminate that overbanked element?
I feel that might be unnecessary. Worst case is they fix this support and add another on each side. My guess is the footers sank a little causing extra load on the joint.
I might be wrong but don’t forget during Winterfest time Fury 325 was running in 2022 in the Winterfest Season.
I believe that it has to 40° or 45° for Fury to run. Most tubing and pipe isn't effected by those temperatures.
The temperature rating ensures the train can clear the circuit.
Being in the tool and die trade, I love SOLIDWORKS!!
Word is a park goer spotted the crack the week before and nothing was done about it.
Very quick and interesting analysis!
Glad you think so!
Everyone calling this a “design flaw” how about you see the support was fully detached and the ride was still running flawlessly. This rides an engineering masterpiece quit over reacting.
It sure is a masterpiece. I didn't call it a design flaw. Apparently, it was designed well so that if one support is severed it didn't fail.
Excellent analysis
Sad thing thing when it was brought to the attention of park officials, they wouldn't shut it down, It took calling the fire dept and they shut it down.
got a link to this claim?
@@sadiporter2966 NUMERUS TV news stations reported on it.
@@sadiporter2966 CHANNEL WCNC video titled Fury roller coaster closed after support beam crack repored.
There’s pictures of the crack starting from over a week and a half ago
Cool. Where?
Give us a run down on how close they were to the tracks failing
I can't say definitively. Hopefully the design factors used by B&M would allow for a few thousand load fatigue cycles (dispatches) before failure.
I've never seen angled supports angled in the same direction as the forces. I'm no engineer but it looks like bad design and made this crack inevitable. They should be angled against the forces.
It would make more sense to have that angled away so the joint would be in compression. However, that would put it in the parking lot.
@MentoredEngineer lol they should have sacrificed the one to three parking spaces it would have taken to place the correct supports
B&M had designed lots of these. This one wasn't unusual. Check out this one on Great Bear! thedod3.com/wp-content/uploads/2013/06/greatbear12.jpg
The post is designed to have some flex, your solution is to remove the flex. That will ultimately result in a catastrophic failure, where in this case, the other supports acted a redundant support, and took the load. Two things will happen. The entire support will be replaced By B&M, possibly more as they took unusual stress for an unknown time period or the ride will be removed. Hopefully the Park President and Maint. Director will be fired for a culture of lax inspections. This was likely the result in three things, not two. Weather as metal expands in the heat, unanticipated stresses over time and what will be eventually stated, an unforeseen failure in the mfg process of the support. The latter being easier to state than design flaw as there are other coaster made similarly by B&M across the country. Cedar Fair will downplay the risks and safety and hope the final story is buried. The Park President and Maint Director will be left untouched as no one was hurt and the park did not get sued.
Cracks on coasters happen more often than you think. Look at the Bat or Big Bad Wolf.
Your opinion on the flex is also incorrect. Military aircraft, like cargo planes, have wings that move 10-12 between on the ground and in the air. Commercial airliners maybe have 1-2 ft of sway under normal operating loads. We don't see airliners falling out of the sky because they are stiff. It is all in your design goals.
There are probably way more than 3 things responsible for the failure and the investigation will find them all including human factors (operations, management and inspections) as well as the mechanics.
The temperatures in question here are not in a range that will effect the stresses in any significant way.
@@MentoredEngineer Please read what I wrote and not what you think I wrote. The post needs flexibility, otherwise it would snap like a toothpick. Your solution appears to remove flex and movement. A stiffer system is unhealthy. Watch every roller coaster, they bend and move. I also attempted imply the weather was a factor, not the total answer. 'Cracks appear more often than you think,' yes, aware but they are usually caught. This one was not. Also, this coaster is a baby at 8 years old. I can tell you stories about what appeared to be catastrophic failures which went unnoticed, and no one got hurt. Right now every Park is re-inspecting everything, including Six Flags, all Cedar Fair, Merlin, Universal and hopefully Disney to name a few. However you are still safer on ANY roller coaster than in the car ride to and from the park. Thanks for the videos.
Ok, great!
Seems like they should have a support that’s “pushing” rather than “pulling”, right? For example, the support looks like this: /|, would it not be better as this: |\ ?
Less chance of failing, but it might have put that support in the parking lot.
As we are talking about fatigue induced failure, putting the support on the other side would have one advantage: the weld toe would be in compression instead of in tension. While the design codes for welded parts only asses the stress range and neglect the direction of stresses and the mean stress (many good reasons to do so), doing fatigue tests on my own, I know that welds under compression perform slightly better than welds under cyclic tensile stresses.
However the whole argument goes down the drain, if weld flaws are involved.
A huge question now is could other B&M coasters using this specific type of support column also be at risk for similar stress cracks forming?
It is doubtful that the fundamental design of the support is the issue. Many are saying that the diagonal tube's footer sunk a little which would add extra load to the joint. B&M has a great track record with this design.
Dude could use his neck as a model of how to engineer a proper support beam.😅
What do you think is a better alternative to a tube to tube joint, since most roller coasters use tube to tube joints?
First off, B&M knows what they are doing and the round tube joint has a proven track record. However, I prefer square tubing because it mates together better. This also allows for the easy addition of doublers and gussets that can help ease the transition. Square tubing does cost more though.
@@MentoredEngineer However the fatigue strength of square hollow section joints is much lower than circular hollow sections, hence you would need thicker cross-sections to compensate.
It seems to me like a laser scanner would be a good investment so these things could be caught early. Super interesting video.
Thanks! I'm not sure if laser scanning is a Non-destructive testing method. Do you have any info on this technique?
A laser scanner probably wouldn't have found anything except maybe the settling of a foundation if that exists and contributed to the failure. The only way to find this earlier is pretty much a manlift to do mag particle testing. That being said this crack didn't go to a full break in one day. They definitely need to improve their visual inspection process to make sure all joints and connections are subject to a close visual inspection (using a zoom lens or drone) probably on at least a weekly basis.
@@danbert8 You can get some pretty good results on integrity with a modern scanner - just not early enough to stop the loss of integrity from happening. I agree - the scanner becomes helpful when its already "too late", but they could have probably caught this months ago and have already mitigated the issue. I just assumed that parks would be scanning footers and such regularly and comparing the model data to make sure there arent issues. I wish I was still working for Trimble - I'd be having a field day with this.
@@MentoredEngineer I'm guessing they're referring to using lasers for very precise measurements to create a 3D model of position. That would, of course, be non-destructive, and any (unexpected/unallowable) change in position would require further investigation to determine why something isn't where it should be.
@@rvplc9428
Lasers can find displacement, not cracks.
I love when the general public trys to argue with me, the amount of people who thinks they just gunna drop a new support in or weld it back up in just a few weeks.
these people have no idea about construction. inspections alone take weeks if not months. and we dont know yet weather they will have to retro fit new supports or even foundations on other parts of the ride. I just hope they do this right
I doubt the weather has anything to do with it. It looks sunny in the video. ;)
Indeed. The investigation will take 2-3 months minimum. B&M and Carowinds will be working on repair while that is going on and then implement it quickly. It may be open for Scarowinds later.
@steveversteegen8217 we not talking about direct weather 😒 we mean temperature expansion and contraction.
It seems B&M is at fault (based on this video on the joint).
I won't necessarily assign blame based on a video from a hundred feet away
@@MentoredEngineer That is a full penetration weld all the way around that joint.
Thanks for clarifying
@@MentoredEngineer, I could absolutely be wrong, though.
Do they re-heat treat that area after they weld it? Seems poorly designed for what it is being asked to do.
My guess would be no. This isn't high strength steel so any type of heat treatment would yield minimal results.
Could be a opportunity to make some more changes also rebuild the whole thing..maybe redesign that whole section..maybe adding a loop:)
They'll want to get it up and running ASAP with a minimal capital expenditure.
How bout another support post from the point of track connection coming off the right hand side down to the ground at angle?
Would not change anything, because the problem is a the point where the column head meets the weld of the diagonal.
Maybe a better explanation: if you mount an antenna on the roof of the building and the anchor points between antenna and roof fail, stengthening the building will not help.
The way that support is designed, there is shear force working at that weld joint. Would it not be better to have the support brace on the outside so that the joint is now in compression when it is stressed?
From the map, it looks like it would be in the parking lot.
@@MentoredEngineer ok, but that wasn't my question.
@@shrodingerschat2258 Sorry: Yes, but that support would be in the parking lot.
Read where the amusement park said they inspect daily (like all amusement parks say).....calling B.S The crack is to large/ long to be just 2023 season problem......
Maybe. It takes a long time for a crack to start, but once it is going it can fail very quickly.
Structures aren't a daily inspection because it would be too time consuming. Woodies only have to have the track walked each day, not the entire structure.
From now on to be known as 'Kraken'
Perfect!
looking at the comments of mainstream reports of this, its looking like we may see short lines when it reopens!
I'll be there opening day!
Your mic is really fuzzy, lots of static.
Noted
Thanks!
Beer money!!!!
Thanks Mike! I love beer money!
It should never have gotten this far and the fact that it did speaks extremely poorly of the safety of the entire park
You lost me at “inadequate design”. You jumped to that conclusion before even considering manufacturing errors, bad welding, inadequate NDT, ground settling, or other non-design-related possible issues. Several engineers with many years of experience spent a lot of time designing this ride, and it had to have been signed off by PE’s in order to get built in the first place. Saying the entire support needs to be redesigned is a HUGE leap. I’m not saying it’s impossible that there was a design error, but all things considered it’s very unlikely.
I know this incident is fresh and is surely getting you clicks, but please consider the possible effects of what you’re saying here. As an engineer you should understand the ethical implications of spreading public mistrust.
Apparently you missed my disclaimer. The load has not changed since the day it opened and it broke; therefore, the design was inadequate. The investigation that will ensue will identify the root causes.
Rest assured, we will learn more from this one broken support than all the rest of the successful supports.
Should had been inspected under load before the was open to public.. ;-(
That would violate so many OSHA regulations. Having a guy inches away from the train as he is inspecting.
Even if that would be allowed, if we are talking about fatigue cracks, those develop over time, hence won't be there when the ride is assembled.
But if you think of checking each column each and every single day before opening, you need staff for that. Or time. Let's see, the track is about 6600 ft long, spans between columns are roughly every 20ft, probably less. Makes 330 columns. How much staff do you want to put on this task? And would you be willing to pay for this staff?
@mentoredengineer can you comment on the next likely occurrence if the crack had not been found?
I know this is just prediction, but my husband and I are big coaster enthusiasts, we just left Hershey this morning! We have a small build in our backyard ❤
Weve discussed, what could have happened if it were continuously running as is. I have no idea, but he seems to think it could have ran a lot longer without any real catastrophic event.
Well, the load was distributed to the nearest supports on each side so there was no immediate danger. However, this increased loading will decrease the lifespan of those supports.
Kudos on the backyard coaster. I'm making a backyard coaster database. Feel free to enter yours. mentoredengineer.com/backyard-roller-coaster-database/
What's the worst that could have happened? Im no engineer or anything.
Well, if the track were to break or even just kink, the train would derail or come to a very abrupt stop. Most likely this would cause injury and/or death.
@@MentoredEngineer Of course makes perfect sense.
@@MentoredEngineerAt least it's nice to know that beam isn't the only support holding up the track, or the only support holding up the train for that matter.
Your analysis ignores a very obvious issue with this design: Expansion in the track and no joints to accommodate it against a rigid support system. Better welds will not eliminate that stress and ultimate mode of failure.
I doubt that adds more than 2-3% of the load. These things are less rigid that you think.
Not even a month later . . LOL
Im a structural welder. I seen the close up photos. Lots of under cut in the top edge of weld. And thats where it broke. That welder shouldn't be on that job.
Wow! Such judgement from a photo taken from over 100' away. Hubris.
Great vid! Maybe they used Gorilla Glue like they did on the Hickory NC arches 😂…jk. Glad that it was noticed and no one was injured.
Amen to that, but JB weld would be the better choice. HaHa
@@MentoredEngineerwhat about Mighty Putty with Billy Mays
Yes, it’s horrible that it’s shut down. But it’s better than having many deaths.
Season ending injury
Oh no... Cedar Fair already hates Intamin..........and RMC.... Now they're gonna hate B&M
LOL - Cedar Fair has some real short sighted people high up.
I dealt with a similar situation where 1980s fiber optics were banned and we had to wait for everyone that was around at the time to retire before reintroducing fiber-optics
This is not the only one that cracks were found at. Most of this steel made today is garbage. Any ride you see closed is a good sign that cracks were found at stress points.
Most manufactures will insist on Mill testing for each lot of steel so they can trace defects like this.
Except that cracking usually does not happen within the base material itself, but at the weld toe or inside the weld.
Having no clue, but having a strong opinion and spreading it. A typical youtube commenter.