Pre-cambering Steel Beams 8

[damo74]

Hi guys. Has anyone got any advice on pre-cambering steel beams?

I was going to pre-camber an amount equal to the total dead load for a beam supporting precast concrete slabs and blockwork walls. Then all my beam should deflect is the live load deflection, which I've calculated as being within specified limits.

Am I on the correct train of logic?

 

[JedClampett]

Not a bad idea for longer members. Refer to sheet 1-147 in the AISC ASD 9th edition for standard practice.
If your camber is less than the minimum they can provide it's not worth the bother.

 

[SlideRuleEra]

Concur with JedClampett. For more background, here is link to a classic article in the Septermber 2004 issue of Modern Steel Construction

http://www.aisc.org/MSCTemplate.cfm...Management/ContentDisplay.cfm&ContentID=26794

http://www.SlideRuleEra.net

 

[tincan]

In the mid 70s Fears Structural Lab did some testing of composit sections constructed by inverting the slab and beam. The inverted beam was supported at the end points and the slab was poured into a frame attached to the beams.
When the slab reached 7-day strength, the beam was inverted. This preloaded the beam such that when the beam was inverted, you had cancellation of the dead loads (almost). Saw the installation of several short span structures using this method.
Best, Tincan

 

[flamby]

Then all my beam should deflect is the live load deflection, which I've calculated as being within specified limits.

NO.

Your camber+LL Deflection should be less than allowable deflection.

Rule#1- Never pre-camber steel beam until absolutely necessary as driven by service requirement.
Rule#2 Never ever pre-camber steel beam for any structural reason.

Ciao.

 

[TPNY]

flamby, are you saying that you subtract your camber from your LL deflection to compare to the allowable deflection? ...because it doesn't make sense to add the absolute camber deflection quantity with the absolute live load deflection quantity-- they are deflections in opposite directions. I disagree. Or, were you talking about total deflection (service DL + service LL)?

 

[WillisV]

Something else to think about in cambering is the economics involved. As summarized from a recent Modern Steel artical it costs about $50-65 dollars per beam to camber a reasonable (0.75"-2.5") amount. With the current cost of steel running about $0.30 per lb this equates to a weight "cost" of 167lb to 217lb to camber a beam. Considering a 30 foot span this would equate to around 5.5-7.5lbs per foot. So to camber a W18x35 verus using an uncambered W21x44 is that would mean the cost differential would not be large (not considering other considerations such as studs etc) to go with the deaper non-cambered beam provided head room etc. was available.

 

[JedClampett]

Willis, you're picking the shortest beam possible to work your numbers. It would be unusual to camber a 30 ft. long W18 beam. If you're talking about a 50 ft. long beam, you're only increasing the cost by an equivalent 4 lbs. per foot. Plus, even though you would get a reduction in deflection, using the W21 vs. the W18, you would still have a deflection that might require a camber.

 

[WillisV]

Agreed just wanted to make sure people thought about $$$.

 

[pba]

Flamby,

I'm not sure I agree with you. Live load deflection is surely the criteria as this is (potentially) cyclic. Deflection due to slab weight is a one way only load - Once it's there it can't be removed.

I'd be careful with significant wall loads in setting the pre-camber value. It might be better to treat these as live load...

 

[flamby]

pba,
(DL+LL)Deflection should be less than allowable deflection. Providing camber will not change the total deflection. I don't think, I can take advantage of camber in how I show the deflection is within limits. I don't think the deflection limits of beams, as specified, care whether the load is deal or live or self.




Ciao.

 

[pba]

Flamby,
You have a different approach to us - we check total deflection or imposed load deflection only, depending on the circumstance. Clearly total load deflecton IS important in lots of cases but imposed load deflection is usually more important as this is the load that could be removed and replaced leading to actual deflections occuring in the building. Dead load deflection is just a curve in the frame. Once it's happened it's permanent. It does not crack walls or cause doors to jam.

 

[Barbicane]

Camber in a roof can cost in other ways. It can divide the water in two and generate a need for more roof drains and plumbing labor. Raising one end of the beam to eliminate the ponding, at least twice the camber, creates a lot of extra detailing costs. Another way to neutralize the water effects of cambering is to get the rofer to install a system of insulation wedges. This leads to having multiple details around the edges, usually a parapet wall. Certain guilty parties, who will not be named, have been known to procure roof replacement without the wedges and ended up putting down a third roof with the wedges. People with electronic equipment and electrical switch gear do not like to look up and see plumbing over their equipment.

 

[Stress02]

I have been told by a personal friend who owned a sizable fabrication shop that it is a gamble when it comes to camber. Many times the camber changes or is lost by the time the piece gets "thrown off of the truck", so his advice was to save the money.

 

[RVQC]

On beams less than 40 foot whats the most effective way to add 1/2" camber. currently using a consumable rod to camber beams, but it is a crap shoot to keep any tolerances, and leaves the beam a mess when all is said and done. Are there better ways to camber smaller beams and stay within 1/8" tolerances

 

[rapt]

Precamber will result in a reduced visual deflection which is one of the requirements of deflection checking.

Precamber will not reduce the deflection experienced by anything supported by the beam. So vibration checks done based on LL deflection and deflections experienced by brittle finishes such as block wills will not be reduced. The deflection is simply measured from a different starting point.

 

[14159]

I have a stupid question: What's the difference between camber and pre-camber?

Be careful with those AISC tables that give camber info. They're MILL CAMBER, not camber put in by the fabricator. The upcoming Manual will have an improved discussion of this. Look in the 3rd Ed. LRFD Manual Code of Standard Practice. More useful information is given there.

I did an experiment a few years ago to test camber accuracy, although admittedly in a statistically-insignificant way. I had 6 W16 and W21, unshored composite beams that were cambered 0.75" using 85% of the DL deflection. They were randomly scattered over a large hospital floor. I talked the inspector into measuring the camber at the following points:
1. before cambering (the natural, or mill camber)
2. after camber
3. just before transport
4. after deliver to the jobsite (a several hundred mile trip)
5. in the air
6. with wet concrete on there

I was amazed at the accuracy and precision for the 6 beams in my little experiment. None were off by more than a 1/16", and the small errors were all under-cambered, which is good IMO. None of the camber came out during transport. They were just about perfectly flat (measured, not eye-balled, although I can't remember the numbers) with the wet concrete on there.

14159