Curved Concrete beam in elevation

[KOTOR1]

I have to design a curved concrete beam in elevation for a retail building. The architect wants it to have a tunnel like feel when entering the building.

The length of the beam in plan is about 35' feet. The curved length of the beam is going to be about 55' long.
Are there any books or code references that I could refer to for concrete curved beams? I am guessing the curved beam would result in a horizontal force at its ends along with the vertical down force. I am having second thoughts about designing the beam, because I think the spans are too big

Would it be much easier,using HSS tubes forming the curve in lieu of the concrete and then infilling w/ wood/steel studs to give the curved appearance.

 

[csd72]

So its a semi circular arch then?

Steel beam with timber framed cladding is always cheaper but not as durable, best to check with architect if this is acceptable.

Look up Roarks or similar for a formula on this, it can be designed with no horizontal restraint at the base.

 

[priya15]

I have used curved HSS tubes before for a design that I did for a walkway. It looks like a space ship landing dock. Anyway, steel hot rolled sections are bent that way a lot of times and it is not unusual. I suggest you specify a curved steel beam. Make sure your connections are designed appropriately. This is the easiest way to handle that problem in my opinion.

 

[ron9876]

Model the beam and supports for boundary conditions. First with roller supports and second with restrained supports. The answer will be somewhere between and if you can design for the worst case and get a reasonable result then you have it. Will only be a question of how big it needs to be. Unless it has very large loads it sounds like it should work in concrete.

 

[csd72]

One more thing, you need to be extra careful about lateral torsional buckling if you use a wide flange.

 

[BAretired]

How will it be formed? Forming it in place would be quite expensive. Maybe you should form it flat on the ground and hoist it in place.

Or you could design it as a three hinged arch and pour one half on top of the other, then raise both halves with a couple of cranes and connect all of the hinges.

A three hinged arch is statically determinate and easy to analyze. It requires horizontal reactions at the supports to support gravity load.

BA

 

[ishvaaag]

I think just modeling the structure in an able 3D program will get you reasonable approximation to the behaviour. If it is curved only in plan, just caring for the forces and displacements should end with the problem, provided you have enough structural clearance for RC construction. If the beam is curved in plan and elevation, I woud try to brace it to some structural elements at the façade, bridging with the canopy roof elements; stabilizing this way laterally the thing will diminish further troubles.

 

[lexim]

Hi,

A good reference book is " Reinforced Concrete Designer's Manual" by Reynolds & Steedman. An excellent book.

Lexim

 

[inju]

Check "Structural Concrete: Theory and Design 4th Edition by M. Nadim Hassoun & Akthem Al Manaseer" It is kind of decent RC reference. I believe it covers design of beams curved in plan but not sure about beams curved in elevation.

 

[kslee1000]

The benefits of HSS lay in its torsional stiffness, and considerably lighter than concrete. But bend to tight curve and splicing by field welding may cause some headaches.

 

[BAretired]

We don't seem to be getting a lot of feedback from the OP on this thread.

BA

 

[kslee1000]

Actually I don't even know if he means curve in horizontal, or vertical. What is "tunnel" like feel? How it is to be supported? A lot of details are missing. I remember there was a similar thread a while ago, but don't remember seeing the end of it.

 

[BAretired]

kslee,

I assumed from the title of the thread that we were dealing with an arch. The dimensions given, 35' span and 55' curve length seemed to be consistent with a circular arch. If that is not right, the OP should let us know.

It is not possible to help people who do not provide adequate information.

BA

 

[hokie66]

I agree that the shape described sounds like a semi-circular arch, but it doesn't have to be designed as an arch. Without horizontal reactions sufficient to resist the arch thrust, it could just be proportioned and reinforced as a beam spanning 55'.

 

[csd72]

yes pi/2 times the rise equals 55' as you would expect for a semi circle.

KOTOR1 are you still there?

 

[KOTOR1]

Sorry guys, I was in Florida.

We just got revised backgrounds from the architect and it seema a lot has changed since I was in Florida. The walls are now revised to ICF walls. The openings size has been increased to 40'(end to end) and now it is a elevated complete semi circular arch supported by 14' tall vertical piers/jambs. Originally, it was about 1' feet short of being a complete semi circle and it was directly supported on the foundation. The curve is in elevation. It looks like this (Elevation View)

/ \
/ \
I ^ I
I I I I

^ - is a semi circle.

The arch is now taking more load since the roof joists are now spaced at 30' (orignally it was 8'). The arch is at a non bearing side of a hip concrete tile roof w/ SIP panels as the roof. The architect wants to use Metwood as the joists. Has any one used this before?

So I can assume it be a straight beam (span = pi/2 * 40). I will look through Roark's book this weekend. I will try to get the other references mentioned also.

 

[BAretired]

So how were things in Florida? I only dream of going to Florida.

If you have an end to end dimension 40', the span is 40', not pi/2 * 40. It is a simple beam, although the loading is not likely to be uniform.

Personally, I don't think that is the way to go, but let's not sweat it too much. The architect may change his mind again.

BA

 

[KOTOR1]

Florida was fine. But, explaining the live load problem(s) as per the other thread to the owner was a real pain.

So I should design the beam as a simple span of 40'. Isnt the curved length its entire length (half a semi circle)?

 

[BAretired]

Kotor,

You can have a beam which undulates up and down, and, if the support is a hinge at one end and a roller at the other, the moment at any point will be precisely the same as a straight beam. That is a fact upon which you can always rely. So the moment at the midspan of your strange shaped beam will be wL[sup]2[/sup]/8 if the load is uniform along the length of the beam. Moreover, that will be the maximum moment in the beam.

If you elect to use a hinge at each end, the structure becomes indeterminate to the first degree and you must consider conditions of compatibility.

If you use a fixed end at each support, the structure becomes indeterminate to the third degree and must be solved by one of the many techniques to which you have no doubt been exposed in your engineering education.

BA

 

[prex]

So you have an arch, not really a beam: other responders above were in doubt, thinking it was a curved beam in a horizontal plane.
BAretired is of course correct in its statement concerning the maximum moment at midspan (oops...at the tip of the arch). However he perhaps didn't fully catch your description with the arch sitting onto two columns: this makes it necessary to consider the arch as pinned-sliding, unless the columns may be made really massive and quite short or you want to embark in calculating the behavior of the complete frame (columns+arch). So, for a vertical load that's uniform on the horizontal projection, the maximum moment is indeed wL[sup]2[/sup]/8, L being the span column to column.
In the first site below you can find calculation sheets for circular arches under Beams -> Curved -> Arches -> Circular -> Pinne-h.slide , with various loading conditions.

prex

http://www.xcalcs.com

: Online engineering calculations

http://www.megamag.it

: Magnetic brakes for fun rides

http://www.levitans.com

: Air bearing pads