The cycles may be cause something to happen.

But you need to identify possible somethings to see if they are actually happening.

A properly designed joint will not lose preload because of cycling loads (within the design envelope).

"why" is nice to know, better to utterly prevent it ... anti-rotation device on the nut, wire-tie the nut, cotter pin, ...

a constant wind in unlikely ... more likely may be a buffet ? or a thermal cycle.

if some outside cause is making the joint relax, then you'd need to physically lock the head and the nut.

another day in paradise, or is paradise one day closer ?

Sounds like you need a lock, this is most likely either thermal or vibration (high freq) related.
Make sure that you are really getting this tight enough and then pin it.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

It seems appropriate to show a photo of the connections. Was this a hot-dipped galvanized structure? Are there countersinks involved? Are you connecting tube? How were the bolts tensioned - per AISC/RCSC? I don't think that properly tensioned fasteners will suffer nut rotation - I would suspect that there has been some sort of relaxation due to your materials moving around.

Unless the wind is causing flutter, it should stay put.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! https://www.youtube.com/watch?v=BKorP55Aqvg
FAQ731-376: Eng-Tips.com Forum Policies forum1529: Translation Assistance for Engineers Entire Forum list http://www.eng-tips.com/forumlist.cfm

The clamp load is not high enough.

Ted

Most car port structures that I see are made of steel tube, is that the case for yours? If you are using through bolts with hollow sections I suspect it is due to the walls of the hollow sections yielding. If that is the case, you need to rework the connections so that you aren't using through bolts, or add sleeves that are welded to the walls of the tube that the bolts pass through.

rb1957, IR, Ed-Yes, I wasn't exactly clear in the OP, but I think the structure design is having some vortex shedding/buffeting. To give some visualization, the roof is basically an inclined flat plate at 7 degrees. So there are possibly vortexes created on the leading and trailing edges, also maybe some downwash or uneven loading at the structure ends. Lock washer/nuts are clearly the next step, but I need to get some numbers to present to my boss and client.

dvd-I haven't mentioned it yet, but I did not design the structure. The only info I have is from engineering drawings, images from Google, and second hand information from my manager. I have attached an image showing one example of the structure. There are different options for mounting the roof beam (cantilever/center mount)... Everything is galvanized steel, no countersinks, and the main members are made from boxed cee's welded together. The drawing shows multiple methods for tightening (torque nut, load indicating washer, and tension controlled bolt), but who knows what the contractor actually did.

Based on the feedback from everyone, I think my new solution is going to be to try and calculate the vortex shedding frequency first. If I have this I can maybe run a modal analysis (and maybe aeroelastic analysis) on the structure to see if there is some correlation. This will probably be the best way to show there is vibration in the structure and then I can just recommend locking hardware.

• https://files.engineering.com/getfile.aspx?folder=09114974-9b2b-4978-bd01-9

if vortex shedding is the issue, how about adding vortex generators ?

ok, that's not really a suggestion, as the previous physical restraints should work against vibration.

if you're saying your need Proof of vibration before doing a simple re-design, then place accelerometers on the roof near (and far) from the fasteners.

but i'd've thought that there's a reasonably straight line of logic from fasteners loosen > vibration loads > physical restraints

another day in paradise, or is paradise one day closer ?

rb-I will hopefully get to the point where I recommend locking hardware, and then tell them to do the accelerometers if they need more proof. Per your last statement, I completely agree. I am assuming someone found the loose hardware and from there the contractors or managers assumed this to be a massive design flaw without giving this an insightful look... Anyways who knows, thanks for the help

I expect its not a vibration issue at all (at least not fast cycles that are associated with vibration and vortex shedding), just the cyclical nature of the load. Your picture shows a cantilever column with a beam cantilevering each direction off the top of the column. I'm sure this joint sees/resists all sorts of back and forth rotation in any wind loading. The bolts can not be properly preloaded because of the flexible walls of the HSS. Locking type nuts or loctite will prevent the nut from loosening to the point of falling off, but they won't help with the clamp loss, as you still have the flexible wall of the HSS problem. As long as your joint performs adequately with a snug tight bearing type connection your locking solution will work. If you need a more rigid connection that requires pre-tensioned bolts the locking devices won't fix that. If this structure is supposed to be designed per IBC/AISC this joint would require pre-tensioned bolts.

dauwerda-I'm not exactly sure if this is what you are saying, but the bolts do not go all the way through the member. There are access holes to get to the interior beams as needed

Ok, my assumption of a through bolt was incorrect. So, flexibility of the HSS sidewall is not a problem. Then yes, I see no issues with providing locking devices to prevent the loosening as you suggest.

Unless it is widespread, it could just be sloppy work / inspection.
If on hollow shapes and they will always work loose, your failure mechanizm is losing the nut and then the bolt - could you just paint the heck out of them?

IFRs-I was told this is an issue at multiple sites, but I don't know how true this is. I have limited info from the client... I am assuming this is a design specific issue at a specific site with a unique wind flow path. Also, no painting allowed since everything is galvanized that removes the need for rust protection and it would add cost. Thanks for the input

I would usually go to Nylok types first, but for outdoor where you have a lot of temp cycle I would use castellated nuts with cotter pins or bend up lock tabs.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

Slipping in a spacer sleeve and running a bolt that passes completely through the beam will do a lot to improve the joint.

I think through bolt and spacer sleeve is a good idea. The wall is too flexible to handle the stress due to cyclic wobbling of the cantilevers. If you have program available, A FEM model analysis can provide better picture on the local stresses.

Might be more like variable wing wind loads than vortices. If you cannot sufficiently increase the clamp load with existing bolts, add more bolts.

Ted

For me, the reportedly out-of-control assembly, and resulting as-yet undefinable boundary conditions would put any analysis efforts on hold.

"First, build it stock." (advice to budding hot rod engine builders before spending big bucks on all the speed parts)
RTFI (the stated take away lesson by one of my College freshman professors after administering one of those tests with written instructions on the first page to "read thru the entire test". Yes, the test questions were brutal. The last "question" was to NOT solve any of the problems but pretend you were working so your fellow students would not catch on.)

Please post detailed pictures of the actual "structure".
Please provide pictures showing the locations on the structure where the fasteners have "loosened."
Please provide nice detailed pictures of the counterfaces of the nuts and bolts that have "loosened." and the surfaces they were attempting to clamp.
Please provide some nice detailed picture of the counterfaces and faying surfaces of the structure's components that were clamped by fasteners that loosened.
As other mentioned coatings etc on the faying surfaces etc can embed and throw away any fastener preload that may have existed on day one.

A lack of triangulation could turn most any outdoor stRuKtuRe into a never sleeping always watching twisty turny version of the much loved ( ) Junker test machine.

After welding sh*t together one of my least favorite former bosses would often say "that's not going anywhere ".
A significant percentage of the time he was flat out wrong because his "designs" were simply flawed. I never pointed that out to him.
Ever since, if I hear any fabricator, tech or millwright say "that's not going anywhere " I can not help but stop breathing for several seconds and bite my lip.

Keeping a nut locked to its bolt will not necessarily prevent slipping and wear of everything in the deficient joint's grip length.
https://www.boltscience.com/images/boltf6.jpg

https://brushdestructor.com/wp-content/uploads/201...

I am nobody but if that solar car port was at my place the 60-70 prevailing winds I get here would turn it into a pretzel. and stuff would get broken loose and fly off.
looking at the structure it appears weak to start with. so the strength and rigidity will have a big factor on those loose hardware. it would work in a very low wind protected area.
would it be possible to revamp or at least rework it strengthen it. put in a windy spot and run the test.

My suspicion is that poor quality-control on the hot-dip galvanized members and the (I think) very short grip length of the fasteners that are employed in the connections results in the zinc coating creeping under load and allowing the bolts to lose tension. The weld flash inside the tube from manufacturing of the tube may also, conveniently, be under at least one nut/washer.

Tmoose, in the tugboat industry we're so disconnected we let the fitters design the piping systems. We only require the use the listed material. Obviously the engineers are lacking in material knowledge if they can't even steer a fitter. We have the stupidest problems.

Tmoose has written exactly what I would have written if I wanted to put in the effort.

JMO but I would suggest following standard root-cause troubleshooting/failure mode analysis to logically understand and resolve the issue rather than guessing about extraneous causes or potential solutions as most in this thread seem to want to do. Work outward from the failure, slowly growing your scope until you resolve the issue. In this instance, a bolted joint relaxed and fasteners fell out so the first logical system to analyze is limited to the bolted joint itself. Given the original question, I would strongly suggest reviewing Bickford's Intro to Bolted Joints and other basic materials before performing a bolted joint analysis to confirm the original design is correct, then confirm with the customer every detail of how they are assembling it. Develop the ol' fishbone and use it to eliminate every potential root cause before increasing your scope.

A note on guessing - do NOT do it. I cannot stress this enough. Its unethical, makes you look EXTREMELY foolish, and ultimately reflects poorly on the rest of the profession. The customer's time and money is valuable, as is your own. If you dont use facts and standard process to draw logical conclusions then you open yourself and your employer to potential lawsuits if your guess is wrong, which it eventually will be. In many instances the best answer is the simplest and you said it yourself...

Quote:

..who knows what the contractor actually did.

Until you rule out bolted joint design failure and joint assembly failures, any discussion of vortex shedding or advanced analysis is pointless bunk.

I believe experience kicks in as well, what works and what doesn't