eccentrically loaded columns 3

[diblazing]

I am a mechanical engineer that is in the process of designing a raised mezzanine structure. The mez is approx. 340" (x axis) x 160" (y axis) and 8' off of the concrete flooring of the plant. Typically I would have no problem with these spec but my customer is requiring only two columns to support the unit. I must place the columns at the center of the 340" (x axis) and 160" apart (along each side). The loads on the structure are in the ball park of 8000 lb Dead Load and 15500 lbs live load. I am not sure what method that I should use to calculate the columns. I first looked into the secant formaula but do not think that it would apply to the worst case live loading moment on one side. The distance from the column that the load is applied seems far to large. I have also looking into beam-columns but am not exactly sure how to apply the equations. I would like to use a pipe columns for the suport and think that a 16" OD with 1/2" wall would handle the loading. I need to be sure though because I can't have a deflection larger than 1/8" at each end in order to protect the product below. Any suggestions are appreciated.

Best Regards

 

[UcfSE]

I suggest you hire a professional structural engineer.

 

[diblazing]

UcfSE,
I appreciate your straight forward response. This may be a little bit above my ME training. I always find it more beneficial to my growth as a professional to first attempt the design before getting it verified or changed by Professional assistance. I do plan on hiring this out for the final design in order to cover my company. This would only be used for preliminary calcs and quotation purposes. However inquiring minds would still like to know.

Best Regards

 

[GregLocock]

I don't really agree about shuffling it off to someone else either. You might as well do some analysis first, then let the guy with the seal look up the answer in the appropriate table.

Design of beams with combined axial loads and moments in the linear range is not very hard. I'd use superposition to get the stresses in the extreme fibres. That's pretty straightforward.

Your proposed solution sounds to me as though local skin buckling might be an issue, and I've got ask, what happens when someone drives a forklift into it and dings the pipe? I'd be very inclined to fill the pipe with concrete to prevent that.

Just eyeballing it I don't think you have a slender column, but Roark has the necessary equations for elastic buckling under combined axial and torque loads.

Incidentally if people are anywhere near this thing, either under it or on top, you'll be looking at a minimum calculated safety factor of 4, and measured 2, in Oz at least.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GregLocock]

Oh, l/r is gettingto the exciting range, so combined with the offset loading gross 'elastic' buckling is more likely.

The stiffness requirement is pretty aggressive, by the way. designing the foundation will be non trivial.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GregLocock]

In fact I'll go further than that and given that you have a flagpole your effective length is rather long, so you are well into elastic buckling. (By this time UcfSE is laughing his head off).

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[dik]

Can you hang it from the roof and reinforce it? If your cantilever is 170" then your column appears to be of the right order...for a freestanding column. You have a significant foundation to support the round column.

Dik

 

[swearingen]

With all due respect, Greg, you are correct in saying that designing the member itself is quite simple using standard materials equations. Also, the dinging issue is a big one (I'm in the industrial field - we see it all of the time).

However, there are other issues that a professional would know to look at: connection details, base plate details, anchorage, earthquake harmonics, etc. Any of these issues could bring this thing crashing to the ground if done inadequately. I concur strongly with UcfSE's suggestion, although I believe a first pass by diblazing is warranted for estimating purposes to get in the ball park.



If you "heard" it on the internet, it's guilty until proven innocent. - DCS

 

[diblazing]

Thank you for taking the time to respond to this thread. I love reading this forum and always find its contents useful. Now that I have you thinking about it I can get a little deeper into the design. Trust me, if I had anything to do with it, the entire design would have four columns with a factor of safety of 6. The main problem is that this is already sold by the owner of my company to these perameters and as the "in house" engineer I get to figure out how to make the airplane in junkyard wars. All joking aside I do fully understand the importance of public safety and proper design. I also agree that I need to have a Structural PE verify the design. On the other hand I didn't become an engineer because I'm afraid of a new puzzle. I do plan on working through all of the preliminaries as best that I can before seeking professional assistance. I find that I learn much more from a project that is my own. Now back to the puzzle! Picture the Industrial assembly like a football field with the only suport columns at the 50 yard line on each side line. From the 30 to the 30 will be a solid structural landing (about 150" in the x and 120" in the y or sideline to sideline for those of you keeping score at home). The remaining distance on each side is an aluminum plank (4' wide) with rails on all side and a hole in the floor. This plank will be hydraulicly actuated from zero (horizontal) to 90 (vertical). I enjoy hydraulics and have that diced and quartered. The reason that I bring it up here is that there will be very little weight on these "wings". It is spec'd at 1200 lb max load or three person limit. As you can see I did say person and had planned on a factor of Safety of four (thanks GregLocock). There will not be any forklift action around the workstand however product will be moved in and out of the area by an overhead crane. The product is extremely fragile and expensive so it will not be anywhere near anything. I had yet to consider the posibility of damaging the columns and agree that it must be considered in the design (i like the concrete idea). I am now in the process of determining worst case loading and believe that it will occur when I have the max possible weight on one side only. I figure that I can then take all of the loading and turn it into a force and moment on the column. This will allow me to do a fairly basic Stress calc to see if I'm even in the ballpark on the column. I then figure that I need to move the load to the outermost fibers of one edge to determine if the support can handle this loading (feel free to set me straight when you read any holes into this). When I move the load to the outer fibers I can also add the load there that the moment will create to have a worst case loading. Is that correct or am I looking to have an axial and torque load seperately? Where can I find these equations? I am also questioning the use of a "flag pole"(pipe) as well. Am I better off designing this with an I beam? My initial thought was that a large round would more equally distribute the load on the outer fibers. I plan on working through the column design. Then working on the upper bolted connection. My customer has "standard proceedures" for the footer and their own engineers to handle that with chemical anchors. Any insight is appreciated.

 

[GregLocock]

As I cycled home last night I decided

1) I'd assume there was only one post taking the live load

2) safety factor of 4 on that

3) I'd be really worried about skin buckling

4) the softening effect on beams with multiple load types was not fully understood in the codes in 1980(I did a paper on this at uni) - it may have been rectified since.

5) a 7 ton crane with a 6 foot reach would look more like a piece of dockyard equipment than a garage's 2 post hoist.







Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[jistre]

Don't have time to really do more than skim, but if I understood the general layout correctly, the load will all be supported at the top of two columns situated along a line drawn through the centers of the longest side.

With this arrangement, you're going to be less stiff than other arrangements might be. Are excessive local vibrations or earthquakes possibly going to make the platform unusable/collapse? What happens to the vibrational response when the "wings" are folded back unevenly?

Batting this around until you get that structural engineer to tackle it is fun.

 

[UcfSE]

lol @ Greg

Dead and live load are the easy parts. You may have to consider seismic loading and response depending on your geograpphic location, which may be all the more important given your support geomtry and deflection criterion. If only it were just looking it up in a table... So for learning, you should dig out the referenced building code and review your options for calculating loads. You'll need to check the vibration characteristics of what you are supporting to make sure you don't have problems during start-up or normal operation.

Really, I think finding the dynamic loads will be the hard part. Other wise you have an upside-down pendulum, a well documented system.

All the design in the world won't mean much when your footing rotates a tiny amount that causes the support beam to deflect more than you're allowed. With a deflection criterion that small, footing rotation may play a big part unless you can have a massive footing or a deep foundation. You'll need to anticipated footing rotation and connection deformation so you can figure out how much of the 1/8" you have left to use for the superstructure.

If you are permitted by your governing code to use it, the AISC specification and manual has information on steel design. I would also consider welded connections to reduce slip and extra deformation.

 

[diblazing]

Very thought provoking! The good news is that I will not be anywhere near an earthquake zone and do not need to factor that into the design. I have also already contacted a PE to esure that the design is safe.
That being said, I have been in contact with my client and they will handle all of the foundation design and liability that goes with that. I am to assume that there is no foundation movement when calculating the overall deflection of the system. After I get through the Static design I understand that I must consider the dynamics of the system and also look into vibration. The wings are to operate in tandem and will only cycle once a day. I also have shock absorbers at the base of each wing to dampen the impact of the wing. This will not effect the speed of the impact to the stop but should minimize the after effects. Now that you understadn the wings. In your exert opinions, will I have a vibration issue? I seems to me that theoretically it will counteract and equal zero on paper. I realize that will not be the real world case but how do I simulate that?

 

[GregLocock]

I spent far more time thinking about this last night. You really need to go and look at a crane with a roughly similar load capacity and span to see how they handle the footings.

As UcfSE says the deflection target is pretty much unachievable, especially with a conventional chemical bolt into a concrete floor. If we up the size of the mast to 24 OD by 1" thick you'd still get 2mm deflection (out of 3 allowed) due to bending in the masts alone. That ignores the contribution from the floor, the bolts, the flange, and the jib.

Design of the joint jib->mast, and mast->base, is difficult, as you will introduce stress raisers and so on which will cause locall skin buckling. It needs to be properly matched impedance wise, which willl use much of your 8' up, if you are going for an efficient design.

Could you use a moving counterweight system to reduce the deflection ? or did someone pull that 1/8" figure out of the air (being polite)?

As to buckling with enormously eccentric loads, Roark leaves that well alone. My guess is Timoshenko has it, I haven't got Timoshenko to hand (must feed the book addiction).

Designing the jib itself will be much easier if you have a lot of height - you'll end up with a structure that looks like this:

http://hawthorn.csse.monash.edu.au/~bunny/knitting/2005oct/westgate_bridge.jpg

See if you can get the deflection in that down to 1mm with a 7 tonne load.

Then investigate the stiffness of the foundation system, and the bolts.

Good project, crazy targets.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[swearingen]

Depending on the code adoption in your state, there's a good chance you are required to design for earthquake regardless of how close you are to the San Andreas or New Madrid fault areas. Granted, the resulting loads may be low, but since wind is not a factor for your structure, an earthquake load combination may control the design.

When you say that you are to assume that there is no foundation movement when calculating deflections, does that mean you were told this is so? Is that part of your design criteria? Or is it a design decision that you have made? As a few have said before, holding an overall 1/8" deflection criteria here is pretty tough.

Greg Locock's concerns about local buckling are valid and can be averted by maintaining a proper diameter/thickness ratio that will ensure the section reaches yield before any local effects rear their ugly heads. For the diameter range you're looking at, standard weight thicknesses or greater will help ensure that no local buckling will occur (d/t of around 30).



If you "heard" it on the internet, it's guilty until proven innocent. - DCS