Beam sized by chart challenged 5

[EngineerofSteel]

My beam size is challenged by reviewing engineer. "check beam size results," he wrote.

I used EnerCalc to size a beam spanning 39.58 feet. The DL is just 90 PLF and the LLr is just 200 PLF.

Using AISC tables, I get the same result: W12x19

But using tables based upon span, I get a heavier beam. Do I need to satisfy both design methods? I have been designing according to DL+LLr and the moment at the connection to the column. Now I am being asked to meet another design standard.

Does the beam have to meet both design methods, or just one. It doesn't seem correct that I should check two or more design methods and size my beam according to the largest beam resulting.

 

[DaveAtkins]

Whoa! Did you check deflection? If this is a simple span, there is 4.26" of deflection (obviously unacceptable).

DaveAtkins

 

[ChipB]

This beam is lightly loaded however, rule of thumb is, half of whatever your beam span is in feet (39.5/2 =19.75), is your beam depth in inches. In this instance, I would start with a W18x35, provided it is laterally braced. I haven't checked it though. Check your deflection. On long spans, that tends to control.

 

[JAE]

You need to use the applicable building code in the area the work is done. The code will most likely reference AISC (in the US) or some other steel specification relative to the native country.

That specification (and the associated design tables, charts, etc.) can be utilized to design the beam.

I 100% agree with Dave and ChipB here - 12" for a 40' span....hmmm. As any good structural engineer knows - you check BOTH strength and serviceability.

 

[271828]

I don't want to sound disrespectful, but is there anyone in your office that you can go to for supervision with this problem? The question seems to indicate that you're in over your head.

I am sorry. I couldn't think of a nicer way to say that.

 

[Lion06]

First, as DaveAtkins points out, the deflection for your beam is over 4" (assuming simple span)!!!
If you check the Sx tables, you do get the W12x19 that you talk about, but that is only designing for moment. You still have to check shear and deflection (and vibrations if applicable).
You can't just pull the size out of the table without checking other criteria.

 

[JAE]

What is your unbraced length, Lb?

 

[EngineerofSteel]

Thanks to all, it is clear. I typically spec W16x36 on buildings with these same dimensions.

However, there are many ag buildings in this area with 12x19 beams, and they have been standing for decades. So, I wanted to pass one through the county office.

-DD

 

[ChipB]

I think a better way of saying what 271828 said is, you need a little bit of hand holding right now, and really, this question indicates your mentor needs to provide more guidance.

I checked the beam on Enercalc. Make sure you have all your parameters entered correctly. Under the uniform load tab, there is a check box to use the beam weight in the calcs. Make sure it is checked. Using 50ksi, I still got the beam was overstressed and the deflection was 4.6" or L/103. I've always kept this to L/360 on an itial design. If I'm checking an existing beam, I'll limit the LL deflection to L/360.

Lb: Unbraced Length. Distance between members framing perpendicular (usually) to your beam. There are other factors which get involved, however, it would be best for you to discover those under the direct guidance of your mentor. No, 271828, I'm not even thinking about Appendix 6. Right now, I haven't gotten my mind around the 13^th edition. Required brace strength and required brace stiffness is blowing my mind.

 

[csd72]

Unlike 271828, I mean to sound disrespectful.

What the heck are you doing using a computer program to design a simply supported beam with a udl. It takes 10 minutes to design this totally by hand, including deflection, stresses and lateral torsional buckling checks (using charts).

 

[ChipB]

Oh yeah, I don't have any Steel Manuals in front of me, but most of us structurals stay out of the first grouping of beam sizes for the beam sections. i.e. W12x16, W12x19, W12x22. These used to be called "junior" beams. W16x36 is a good choice. It gets you out of that classification. For some reason, architects seem to love them.

csd72: It takes 30secs to do it in Enercalc. However, it would be good to do it by hand for a bit until you can realize what the computer is spitting out has an error in it. Enercalc should put deflection in yellow if it is <L/360, and red if it is < L/240

 

[271828]

Geez, ChipB, I think your version sounds worse!!!

 

[271828]

csd72, I'd probably use the program too. Mainly because I can save the file and then modify it later when the architect makes it longer by 2'.

I agree it's always prudent to run a manual calc when there's any question about the program.

BTW, ChipB, what's up with avoiding W10 & W12? I wish I had a quarter for every one of those I've used. Not 40' long of course, though.

 

[JLNJ]

Check Cb in the enercalc program. Some programs default to a rather liberal value instead of using 1.00. It gets on my nerves.

 

[ChipB]

271828:
It sounded worse?? Crap.

Nothing wrong with W10 or W12
Most of us have favorite sections. I have one with almost every group from W8 to W24, but W12x26 is my section of choice.

BTW, technically, you probably do have a quarter for every one you have used.

I did get a 13th Ed. Steel Manual from a fabricator. That was cool. I'm wondering if that would add up to a quarter for every one.

Also, do you understand this brace strength and stiffness? I tried to talk to the lecturer yesterday at the seminar, and his direct words were, "Good Luck"

 

[271828]

ChipB, yeah, it sounded worse to me. You ogre, LOL.

You're probably right about the quarters.

Understand brace strength and stiffness? That's a really funny one. I've spent A LOT of hours with that stuff and I think I've scoured every paper on the subject.

The answer: "NO," well maybe "Yes, but only for simple cases" would be better.

I think the problem is that the equations were made up for very simple cases and these don't translate worth a darn to general cases. It doesn't help that lots of the seminar examples are confusing and poorly chosen.

It also doesn't help that the real equations are in the 13th Ed. Commentary and the simplified versions (that'll get your brace killed lots of times if you try to use them...) are in the Spec.

Some of it is completely counter-intuitive also, to the point that I doubt the validity. For example, say you have a beam and are applying the lateral bracing stiffness equation. The more braces you have, the stiffer they all have to be. That's because it must buckle into a sinusoid with a shorter wavelength which is harder to do.

I was in a very famous steel professor's office not all that long ago and he was contemplating teaching that stuff in class. He looked like he was about to give up.

I'm kinda hoping somebody will take it upon themselves to re-invent this stuff. It can't be THAT hard.

Maybe we should build shell models of beams + braces, make the beam initial geometry out-of-plumb, then solve directly for the buckling load. That'll be in the 15th Ed. Spec., LOL.

 

[ChipB]

Solomon and Johnson,"Design of Steel Structures" discussed bracing beams used for bracing need to be able to withstand 2% of compressive force in the beam they are bracing. They also stated this was conservative. If I remember correctly. It's been a few years since I read it. However, I did calculate the compressive force as they did, for the maximum allowable moment in the beam, not actual, and selected beams that could withstand 2% of that force in axial, and whatever bending it may have due to loading. That was long before the 13th Edition was out.

 

[271828]

Yeah, the 2% thing has been around a while. I think its origin is the same as the bracing provisions--a Winter paper from way back?

An example of the problem I have with the current provisions is a beam-column. Does one compute the stiffness and force reqd for axial and flexure, then add them? Linear interaction? Squared interaction? One or the other?

Another example is a rigid frame beam conn to a composite slab. The bottom is unbraced now and they say we can't use IP for braced points. Try using the slab as a torsional brace and the web distortion term eats your lunch.

Yet anothe example is using a shear tab coming into the side of a girder as a girder torsional brace. The web distortion depth should be just the distance from the shear tab to the flange, but the Spec. equations clearly show it as the entire h and it kills your brace if you try to do it that way.

Anybody reading this is bored by now, so I'll get off my soap box!

 

[dik]

csd72... 10 minutes? only if it takes 8 minutes to find a steel book!

Mf = ql^2/8... and 0.000624MsL^2/I for delta

Dik

 

[COEngineeer]

Some of us work in residential and W10 is mostly what I use because you can hide it in joist space (2x12 or 11 7/8 I-joists).