Point load on Water Tower Shell 2

[RFreund]

How would you determine the allowable point load on a water tower shell?

1.) Ask a manufacturer (do these exist?)
2.) FEA shell analysis or something along these lines?

Thanks in advance!

EIT

http://www.HowToEngineer.com

 

[ishvaaag]

If you are facing the problem it is in your interest to get to your own value which normally would be attained by a proper model in a FEM program and of course engineering judgement involving codes and science of construction.

 

[RFreund]

hmmm... yeah i figured. I would really like to see an example of how that is done.
Currently I don't need to perform the analysis (not in our scope of work) but I could see needing to do this in the future.

EIT

http://www.HowToEngineer.com

 

[JedClampett]

Preferably, I would get the original designer or another tank design firm involved. Otherwise, if someone put a gun to my head, I would add another plate of a size that the current plates stresses are insignificant for the new load. I would be worried that if there was damage in the future, you'd be on the hook for that and anything else that happens to the tank.
Also, I'd be worried about welding to a tank. Any coatings, inside and out, will be damaged by welding and have to be repaired.
Our company has turned down this type of work in the past. Cellular phone companies pay big money to attach their stuff to water towers. And the municipalities are very anxious to harvest any dollar they can. But we figured our risk wasn't worth the (small) fee.

 

[JStephen]

It may be possible to address local loadings by WRC-297, WRC-107, or similar references. Some of the vessel design handbooks have simplified methods.

 

[RFreund]

JedClamp - That is basically the situation except the loads have been less than the previous design. The previous design calcs show an "allowable reaction" that just magically appears or I am missing that part of the calc.

JStephen- thanks for the reference.

EIT

http://www.HowToEngineer.com

 

[RFreund]

I am still curious about this. Is there a simplified way or even a complex way to analyze a water tower subject to external point loads. Or put another way - Is there a way to analyze a steel 'domed' shaped shell subject to one or many point loads. I'm really interested in a simplified approximate method but I'd be open to complex analysis tool as well.

Jstephen - I started looking for WRC-297 and WRC-107 what are these and where can I find them. I have started searching and have found numerous threads on them but I'm not exactly sure what they are, let alone where to look in them.

Thanks again!

EIT

http://www.HowToEngineer.com

 

[JStephen]

WRC 537 appears to be the current implementation of WRC 107, available here:

http://www.forengineers.org/store.php?crn=55&rn=799&action=show_detail

WRC 297:

http://www.forengineers.org/store.php?crn=65&rn=542&action=show_detail

The basic issue is analyzing pressure vessel shells for external loads applied at piping nozzles. So the shells in question may be cylindrical or F&D heads (treated as spherical), and the attachment may be a pipe or a rigid plug. Assuming some sort of reinforcing pad is used, you then get back to the rigid-plug analysis for a point load. For the full-blown case where you have 6 loads (3 forces, 3 moments) with a nozzle, you wind up doing a lot of graph-reading and interpolating using the older versions. I haven't used he WRC 537, and don't know how easy it is to use. With a water tower, you may run into issues with the t/R ratios being outside the range covered- pressure vessels are usually thicker relative to diameter than a water tower. The data all relate to either cylinders or spheres, and if youj're actually dealing with the knuckle area, wouldn't be strictly applicable. The loading covered is just from external loads, and the stresses due to internal pressure are not covered.

There are pressure vessel design handbooks by Bednar and Megyesy that have simplified versions of this analysis. There is another common pressure vessel handbook by Moss, which I have not used, and I don't know if it includes similar data. Some of the pressure vessel design software has the WRC 107 analysis built into it as well.

The local stresses allowed for external loads are usually fairly high- in excess of yield- assuming that in most cases, the loads are self-limiting, and due to thermal expansion/contraction, etc.

 

[RFreund]

J - Thanks for the info, I appreciate your thoroughness.

EIT

http://www.HowToEngineer.com

 

[TLHS]

I've done this before. I basically did a FE analysis of the whole thing and used a few different codes to try and set a reasonable stress limit. I can tell you that I didn't find a evaluation system that completely covered the situation. The pressure vessel stuff doesn't quite line up with open vessels, and a lot of the non-pressure vessel stuff is pretty prescriptive and seems to get rather handwavy when you get to point loads or local stresses. In my case it was mostly examining existing items, so there was a greater comfort level that the connected items worked given the history.

 

[RFreund]

TLHS - I think I would be in a similar boat i.e. existing conditions. What program did you use to model it (I wonder if I could set this up in RISA 3D)? What sort of allowable stress did you determine.

I've seen some calculations come across my desk that give what appears to be an 'allowable point load' in range of 2.5k-3k, but I'm not sure how they derived that or if they are even referencing the shell material.

I'm wondering if I could even use a handbook formula from Rorak or Pilkey for point load on a circular plate. But there are many questions as to what to assume for the diameter and end conditions.
I'm thinking if I could create a model; put some point loads on it; see what stress there are; then try to correlate this to a 'handbook' formula, it may be of some use.

The only problem is - with so much 'approximation' when is this useful. Meaning by engineering judgement and experiences you can say < 3 kips is probably ok for 0.25" thick steel. If its 6 kips well now you may be pushing things. But at the same time I'd "like" to go through the exercise.

EIT

http://www.HowToEngineer.com

 

[MiketheEngineer]

How much load?? A 250 fat boy would not scare me - after that I would do a lot of research!!

 

[RFreund]

Yes, research is what I'm after.
I think I may need to crack open Timeshanko's Theory of Plates and Shells and start learning how to model plate elements.

TLHS - I'm curious to know more about how you modeled this or atleast the program you used. Did you use many flat rectangular plates?

EIT

http://www.HowToEngineer.com

 

[dik]

The API has several publications/standards on Low Pressure Storage tanks... LPS tanks. I've been involved with a couple, but for storage of anhydrous ammonia... bigger risk.

Dik

 

[TLHS]

My first step was referencing API 650, even though my case wasn't flat bottomed.

I had an overall system FEM plate model using STAAD, which wasn't brilliant for localized stresses. I did some test cases of partial shell high element count models to see the stresses locally. I don't really trust STAAD that much for this kind of thing, though, so to confirm I did some hand checks of block shear failures and local bending on some roughly equivalent flat plates and also using a method for local loads on pipes by M.W. Kellog (the WRC stuff would fill this same general void).

At the end of the day, I believe I used the allowables in API 650 plus some general engineering judgment to come up with an allowable local stress that I was comfortable with.

 

[TLHS]

Note that mine was petrochemical related, hence the API standard.

 

[TLHS]

When in doubt a reinforcing pad, a straight channel running up and down the cylinder or cone (assuming it's standing on the flat part), a curved beam or some gussets can spread your load pretty easily. Depending on the situation some of these are easy retrofits.

 

[dik]

I realise that the API is not part of this... but, they have a lot of useful information and details for more critical stuff...

Dik

 

[amec2004]

Do a FEM analysis to get the result

anchor bolt design per ACI 318-11 crane beam design

http://www.civilbay.com

 

[VoyageofDiscovery]

I tend to agree with Jed.

If you do venture down the analysis route, a word of caution with FEM, have an actual look at the tower. I once spoke with an old engineer who used to work at one of those outfits in his early years, he recalled some being designed as bladders, not shells. A weak point may be at the cradle supports of the bladder.

HTH

VoD