Thin concrete shell roof 3

[lsmfse]

I have an owner who wants to suspend equipment from a thin shell concrete roof (Monoclastic shell). The shell varies from 4" to 2 1/2" thick. The arch is 80 feet wide. Knowing I can't get proper embedment with an epoxy anchor, I would penetrate through the shell and anchor from the top surface. The other option is to design a ground mounted support system. Any thoughts on this?

 

[MIStructE_IRE]

Thin shells are great for symmetrical and uniform loads. Can you justify unsymmetrical point loads on the shell?

Ground supported would be my preference.

 

[lsmfse]

That's just it. Over the years, the building is almost 60 years old with existing conveyors/equipment supported from the roof. Additionally, HVAC and plumbing upgrades poked more and irregular holes in the roof. It looks like someone took a hammer and beat an opening for the pipe penetrations. My gut feeling without doing some rigorous analysis is to not add any more load to the roof. In full disclosure, my experience with shell structures is limited and I wanted to know if the there was someone on the forum well versed in shell structures. Attached is a picture of one of the post construction penetrations.

 

[JStephen]

My concern would be that if it fails, that failure could possibly be dramatic and complete, rather than a case of excessive sag or something like that. And secondly, if 99 other people poke holes in it and nothing happens and you poke a hole and it collapses, guess who gets blamed?

 

[Shu Jiang PEng SE]

The shell seems extreme thin per the dimension of the arc. Is this the right dimension?

—————————————————————
J&J Structural Consulting Inc.
Structural design, analysis, inspection, drawing review and stamping, and connection design

 

[lsmfse]

That is correct, 80 feet on the radius and 35 feet between ribs.

 

[BAretired]

Ground support for sure.

BA

 

[KootK]

Before we reject this:

1) How much new load are we talking about? 5 lbs? 5000 lbs?

2) Where about on the arc will the load be positioned?

3) Are you able to explain the thickness variation (4"/2.5")

 

[lsmfse]

KootK, the first 8' of the arch is 4" thick and transitions to 2 1/2" thick. We are talking about probably 20 pairs of supports, about 8' o.c. supporting about 250 pounds per support. This is a conveyor as well, so there is some vibratory loading. The supports would be located at approximately at the 1/3 span and mid-span. Based on my limited knowledge of thin shell structures and my research, I feel like the arch action can support additional loading. Although not part of our contract, I did create an preliminary FEA model and the results were favorable; however, I didn't include all the existing equipment loads, I just wanted a baseline. JStephen and BAretired responses are my gut feelings as well.
I've attached a detail that was part of a remodel 45 years ago. The owners rep called us in because the management didn't like the ground supported solution due to the obstructions of the work area. Owners rep needed an independent engineers opinion to back up their opinion. Sorry to be so long winded, but when I traveled to the site, I had no idea this was a thin concrete shell. I appreciate all your opinions.

 

[KootK]

Is there not a way to arrange matters such that you're supporting the equipment from these rib things?

 

[MIStructE_IRE]

Wait..there’s ribs?

 

[Agent666]

I'm with Kootk, span whatever you need either above or below the shell between the ribs. If its 60 years old it seems you have 2 options, either it was designed by some genius in shell design who understood what they were doing, or someone who perhaps had no idea what they were doing but thought they did.

I'd just go the 'easy' route of isolating the loads imposed to the ribs as it obviously predates any sort of modern analysis and if you were to look at in detail it seems there is a lot of unknowns (original design assumptions, additional loading added over the years, etc). If the loads you are applying are small and within the bounds of what has already been added there may obviously be some merit in looking at it in more detail (with corresponding fee to do so) as may ultimately pay for itself if you can save some money in connecting it direct to the shell. Though fixing to 2.5" of concrete really directs you back to fixing to the ribs in my mind.

 

[Shu Jiang PEng SE]

For your information.

An 80 feet dimeter concrete tube with 4” thick wall. Investigate the maximum external buckling pressure of the tube can take. The maximum buckling pressure shall be q=3EI/R^3 (strength of materials: part II Advanced theory and problems – by S. Timoshenko).
Assume the tube length is 1 feet, concrete compressive strength 5000psi.
E=57000sqrt(5000)=4030000psi
I=1/12x12x4^3=64in^4
R=80/2x12=480in
External pressure causing buckling q=3x4030000x64/480^3=7lb/in=84lb/ft. concrete self weight 4/12x1x1x150=50lb/ft.


—————————————————————
J&J Structural Consulting Inc.
Structural design, analysis, inspection, drawing review and stamping, and connection design

 

[BAretired]

If there are ribs at 35' centers, it could be as shown in the first sketch below:

In that case, the shell is spanning in two directions, acting as a barrel vault parallel to the axis of the cylinder and as an arch in the orthogonal direction. The 4" thickening could be intended to act as a beam spanning 35' and carrying the thrust of the arch to the supports.

BA

 

[lsmfse]

BA, that is exactly what it is, a short span barrel vault.
I want to thank all of you for the insightful and thoughtful posts. I get a call from this really good client to investigate the possibility of supporting the conveyor from the roof, and I think "no big deal" (Project is 3 hour flight from my office) The plant manager initially said it was a 12" thick concrete slab. I soon realize it's this thin concrete shell and I completely go in conservative mode. I finished my report today and mentioned that if they want to investigate further, a more rigorous analysis would be required. I had one of my SE's who is a whiz at writing and grammar and who also teaches part time at our local university review my report. She then mentions one of her colleagues at the university is an expert in these type of structures. So now, if my client wishes to pursue this further, I will be hiring this gentleman to do the analysis. They will have to weigh the costs between the analysis, roof repairs at the suspended supports,etc. vs a ground supported bridge type of structure. Again, thanks for all you input. I now know a very tiny bit more about shell structures and am too old to want to know more! I'll post on their decision.

 

[saikee119]

Just a few comments off an old timer:

(1) For a 60 year life the reinforced concrete structure would have been well carbonated. The rebar may not be in the design condition as corrosion could have reduced its effectiveness in strength and bond. To ascertain the strength of the existing concrete is a challenge as one can't get decent core samples off a 2.5" thick shell roof.

(2) The proposed loading whatever it is may not necessarily the worst case as executing the work, say making the roof watertight, could necessitate construction traffic and equipment load the shell has not been designed for.

(3) By the supplied photo the 2.5" thick shell appears to be lightly reinforced, say one layer of mesh reinforcement, so the shell could have some bending strength against flexure. For this thickness it is unlikely for it to be doubly reinforced. A reinforced concrete section with rebar at the mid depth isn't going to offer any significant capacity to carry load.

(4) Due to its age and its slenderness personally I could not see this roof could be structurally modified to satisfy any code or regulatory agency requirements. Also without a statutory approval of a local Government Building Department the unwarranted modification to this roof could void the building insurance cover if it ever collapses.

(5) The modelling of the roof by FE analysis or any thin shell theory is often on the assumptions of an ideal structure perfectly supported, materials homogeneous able to distribute the stress evenly. The structure is a reinforced concrete shell so it will have some minor imperfections from the original construction, weak spots from existing penetrations, loads already added, degradation of material strengths over time and various considerations. These have to be factored into the analysis in order accurately to reflect its true state. This can be very challenging and nearly impossible to verify.

(6) The shell structure could possibly strengthened before adding the new loads. This could be done by a process call guniting of spraying concrete and adding extra mesh reinforcement to build up the structural thickness, assuming the support and the foundation have the adequate capacity. However this route is too expensive for what the required modifications worth.

 

[lsmfse]

saikee,
Excellent observations and comments. I did speak to an SE today highly experienced in shell design and got some great insight. Evidently, the ribs are the primary support and the shell supports the shell as he put it. He mentioned these types of shells are very forgiving, holes aren't really a problem, but the age, vibration of a conveyor and other factors should be evaluated. I considered all your points except your insurance issue in item #4. Good call!. Item #6 isn't feasible for this project. Again, thanks for all your responses, I think the client will accept the floor supported system once they understand the costs associated with an analysis, including petrographic analysis of the concrete, FEM analysis, data gathering of existing loads, etc, etc, etc.

 

[lsmfse]

By the way, STARS for everyone!

 

[saikee119]

Some pointers if you have to do a strength evaluation of the structure.

Petrographic analysis only gives the matrix structure of the concrete. It will give you the information if ASR is affecting the concrete. My guess is ASR should not be a risk in your case as the roof should have a waterproof membrane otherwise it should have collapsed after 60 years due the the 2.5" thickness. Being a roof it should be permanently dry so ASR damage is minimal if any.

The most reliable test is to cut cores. You can select location on the support area to cut say 1 to 3 cores of 3" diameter (need to be at least 3" long after trimming). The cores can give you density, water absorption, ultrasonic (or pulse velocity) and finally the strength of the concrete at the same sage and same mix as the roof. The core holes can be made good by filling proprietary concrete. ACI does permit in-situ strength can be 0.85 of f'c due to difficulty of placing and curing. 60 years old concrete should have strength significantly better than the specified strength due to age effect which does not necessarily apply to modern concrete. Bad concrete will have lower density, lower pulse velocity and poor strength. The three always go together from my experience over the years. If the structure is suspected of under strength a minimum of three cores will be needed to form a judgement.

I would suggest carbonation test on the coring locations as you should be able to see how deep is the carbonation. This is done with spraying an indicator such as phenolphthalein. There is quite a few tests you can do on the remains of the core samples like cement contents, pH value, chemical tests like chlorides which would be recommended if the structure is anywhere within 1 km from the coastline.

The expose rebar size should be measured. Using a cover meter you should be able to get bar spacing to verify any rebar information available.

The above tests are really to demonstrate that you have investigated as far as you can the current structural condition of the structure. They are then, together with the theoretical analysis and calculation, form the basis for your recommendation of any modification/repair/strengthening to meet your client's brief.

 

[saikee119]

Since you do play with dynamic loads in this case I would suggest to find out the frequencies of the disturbing forces and check them against the natural frequencies of your shell roof. . The natural frequencies are the Eigenvalues of the stiffness matrix of your structural system and the mode shapes are the corresponding Eigenvectors. Only the first or lowest 6 to 10 modes are significant to warrant your attention.

In your thin shell analysis with computer you should be able to ask for the natural frequencies of the structure (depends on how it is supported, weight, dimensions and stiffness (E-value of concrete)). As a minimum you should ensure the disturbing frequencies at least 25%, between 0.85 to 1.25 band, away from the primary natural frequencies of the structure.

The purpose of the dynamic analysis is to demonstrate that the new vibratory loads will not self-amplify in the structure.