Properly Sizing Beam & Columns For A Simple Fabrication 3

[Adalius]

When bidding small fabrications, I often am asked to quote simple beams, usually with a single centered point load or two point loads (both symmetrical and asymmetrical), sometimes with a column on each end (I haven't seen us do any jobs with a midspan column). Conceptually, they usually are similar to a gantry. We send all this out to a 3rd party engineer once we actually get the job, but to bid it I often need to get a rough idea on beam/column sizes before getting the engineer's billing meter started. One of our other estimators with no real clue of anything engineering related tried to guesstimate what size beam he needed, added a small margin on the price for error, got the job, the engineer came back with a substantially bigger beam, and he ran over cost and lost money on the job. As a new estimator, I'm trying to avoid getting into that pitfall by using a little knowledge to get me closer to the ball when I bid so I'm not losing money right out the gate...

I have a copy of AISC Steel Construction Manual (13th Ed). Looking at it I understand the concept of moment of inertia (specifically that in terms of beams it can be x-x or y-y for strong or weak way, in general a bigger moment supports more load, that kind of stuff). I also understand that I should be using span / 200 or span / 400 (depending on who I asked) to calculate my maximum allowable deflection in the beam. I'm completely dumbfounded how to relate the information I have into using the available moment vs unbraced span length graphs to properly size a beam (or how available moment in kip-ft relates to point loads for that matter).

I also have the Beam Boy software, so in trying to figure this out, I've used that to sort of brute force find a beam by repeatedly checking until my deflection was good and then going to AISC tables to find the lightest beam I could with a comparable moment of inertia, but I would presume it would be faster if I could just learn how to do the proper calculations. Beam Boy also throws moment and bending stress charts at me, I'm not sure how to relate those against the beam charts to ensure that it will be close to the right size? It will do me no good to pick a beam on deflection only to find out that it has a stress way beyond the steel's capacity or something, right?

So my questions:

1. Am I correct in presuming that when the software throws out a maximum bending stress, I need to keep that below (with a safety margin) the Fy listed for the steel in AISC? I.e. if it shows a 36ksi beam, I need that stress number to be below 30ksi (or something) for example?

2. The software throws out moment in lb-ft, should I be looking at a particular value in the AISC book for a given beam size to compare that against? Do I literally just divide it by 1000 to get kips-ft and use that with the moment vs unbraced length chart? If so do I look up the moment on the ASD or the LRFD side?

3. Given that I have load weights and locations and a maximum deflection, can someone walk me through the steps to utilize that coupled with the AISC book to properly size the beam?

4. Similarly, how do I then translate that into properly sizing columns using the AISC book?

Any help would be appreciated.

 

[BUGGAR]

If you can get the book, Simplified Mechanics and Strength of Materials" by Parker, you can lose the Beam Boy software. Learn what is happening before you let some software do it.
There's lots of stuff you have to know in addition to the basics. Lateral torsional buckling, web crippling, etc. Check how much an engineer would charge for this. Ask why.

 

[CANPRO]

I'm sure someone will chime in here with a meaningful response to your questions...but this is my advice:

You're still going to go to the 3rd party engineer for the final design - so, the more successful you are getting work and completing the job within budget, the more work you will bring them. If you always deal with the same 3rd party engineer, I would suggest calling him/her and explaining you want to know more about sizing the beams/columns for the purpose of bidding the work (with the understanding that the final design still goes to them). If I was in their shoes and you were a regular client I would absolutely do this - because it makes the job of the 3rd party engineer easier if you come to them with a realistic starting point. You will get some good answers here no doubt, but sitting down with a structural engineer in person (especially the one doing the final design) for an hour or two and picking his/her brain will be much more useful to you than answers from the internet. Good luck!

 

[Adalius]

BUGGAR - I have read some statics books in the past, so I do understand a bit more than I let on, but not enough to feel competent. I just ordered the book you suggested on Amazon so I'll be reading that as soon as I see it, thanks!

CANPRO - Well, honestly, it's not that I haven't tried that route (asking the engineer). The story there is the engineer we use is a friend of the bosses from his college days. Nobody at our office, other than the boss, likes working with him, but friendship outweighs business in this instance I guess because nothings changed. So, me being a little peon and not knowing this at the time, called him up, explained the situation, and asked if he could give me some knowledge or a simple set of instructions. The response I got was something akin to asking David Copperfield to give away the secret of his illusion. At the time, I was amazed that he'd respond to a customer like that, but as I found out later and indicated above, evidently nobody in our shop likes working with him (hence why the one estimator tried to guesstimate a beam size). I'd love to change that situation, but I don't have the clout or leverage and nobody else seems to want to rock the boat, so we're kind of stuck.

I was a millwright in the field doing welding and stuff for years too, so I do have some interest in this kind of stuff myself which doesn't hurt. I don't know that I have the desire to go full bore learning 9000 equations for various situations, but given that this design is super common for us I figured what the hell, I'm willing to learn the chunk I need to that's applicable to this somewhat isolated design.

 

[SlideRuleEra]

See Rules of Thumb For Steel Design published in AISC's February 2000, issue of their "Modern Steel Construction" magazine. Most of the information is somewhat dated, it is for 36 KSI steel, but still useful.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[Once20036]

To answer some of your questions:
1) I would avoid anything dealing with stress. Comparing the stress to Fy doesn't account for the unbraced length.
2) Your conversion is correct. Take the unbraced length as the full length of the beam, and use the ASD charts. Make sure that you're showing zero moment at each end of the beam.
3) The "BEAM" section in AISC has a bunch of deflection equations at the end. Find the appropriate load diagram, and use the deflection equation to solve for I. Make sure the beam you pick has an I value larger than your calculations show.
4) The AISC manual has charts for the allowable axial load vs unbraced height of the column.

The above information contains several conservative assumptions that may make your beam estimate a little heavy.
It will be important to get the engineer involved to verify these assumptions and possibly select a smaller beam size.

Go into a couple old jobs. Guess at the beam size with your method, try using the above information to reach a beam size more efficiently, and then compare both answers to what the engineer ended up requiring to get a feel as to whether or not this methodology works for you.

 

[BUGGAR]

And get the free AISC solved example problems off their internet site. I'm using them right now for some design guidance (torsion + bending in my case).

 

[EngineerEIT]

Since you have the AISC manual, I would learn how to use Table 3-10 in the beam section of the manual.

I would use the entire beam as the "Unbraced Length" for the beam. If you are not multiplying your load by a factor, then you are using "ASD" design and should pull the beam strength from the Mn/Omega column. Divide the lb-ft moment you are calculating by 1000 to convert to kip-ft. Then find the lightest shape (represented by a solid line) from Table 3-10 that has a higher strength (Mn/Omega for ASD) than the moment you are calculating from the applied load.

This should give you a good estimation of the beam size based on gravity loading for strength design.

You can also download the AISC Design Examples v14.2 (free download from AISC) and they go through how to use Table 3-10 in their beam design examples. Example F.1-3A should be useful to you for this. I would use Cb = 1 and again use the full length as the unbraced length.

 

[Adalius]

Thanks for the help everyone. Once20036's comments made me realize I was going about this a little wrong by trying to do a bunch of more-complicated things than I needed to.

I solved the deflection calculation for I (in both the concentrated load at center and two equal loads symmetrically placed), replacing the max deflection value with L/400, and wrote it into a function on my graphing calculator. Now I can just plug in the load weight, span length, and if symmetrically loaded, the distance from the end, and the function spits out the minimum moment of inertia needed, then I just look that up in the table like normal. I checked that against a few the engineer did as EngineerEIT suggested and it actually gets me spot on with the 4 I checked.

Does anybody see any problems with this I may have overlooked? Bearing in mind that I'm just trying to get an educated guess on the size since it goes for review after.

 

[Once20036]

The big problem that I see with the sequence you describe is that you're not accounting for the strength of the beam, just the deflection.
You can have a beam that's strong with excessive deflection, or a beam that's too weak with acceptable deflection. While this second scenario is uncommon, I recommend incorporating a strength check into your work flow.

To reiterate, this is a simplified procedure with several assumptions that I don't believe you understand. Always make sure that an engineer is involved before building anything.

 

[Adalius]

Once20036,
I would concur, there's definitely some (lots of?) stuff I'm not understanding Have no fear, they do go to the engineer before production begins (heck, before material is even ordered since we need his sizing). This is strictly a means of estimating a reasonably close value without getting into the nitty gritty of finding the exact.

My thought process is that if my proposed method spits out a W10x15# and it actually ends up being a W10x19#, I can live with that. If it spits out a W10x15 and the engineer kicks back a W18x40#, then that's an issue...

In terms of checking the strength of the beam, is there a particular formula I should be utilizing to determine that, given the information presented?


BUGGAR, that book arrives Wednesday. Thanks again for the recommendation.

 

[cliff234]

Why not get your engineer's "billing meter" started now? That way you don't have to waste your time and your employer's money (paying for your time) - and you can get the right sizes. Ask your engineer to give you quick sizes for estimating purposes. They can probably do it faster and more accurately than you. Ask if they can give you a two or three hour effort coming up with approximate sizes for pricing only. If a friend had a toothache, I could probably remove the tooth for him - but would tell him to go to a dentist. Respectfully, I think you are taking on a task that that would be better performed by an experienced structural engineer.

 

[Lomarandil]

Guys, let's not get ahead of ourselves here.

Yes, there is value in engaging a structural engineer pre-bid.. But there's also cost, and risk that you won't win the job to begin with.

A lot of the decision should be based on the scope of the job -- if you'll be sizing hundreds of these beams, there's more value in getting the bid right. If it's a handful, a few pounds of steel here or there shouldn't break the bid. It sounds like Adalius is dealing with the latter.

As far as stress goes, I'd just compare the general dimensions of beams your other estimates are giving you to successful jobs in the past. Are all of your previous beams short and stout, with big beefy flanges and 5/8" webs? Don't pick a W24x62 this time.

I suppose what I'm trying to say, is there is value in a quick tool to size beams for estimate. For something like a gantry, combining a deflection check and previous experience (like comparing to other jobs) makes a lot of sense.

 

[Adalius]

Lomarandil, exactly my sentiments, most of these are one-offs, and small jobs, but that has a big effect. When a job is $3000, being off a couple hundred dollars due to steel costs is a couple percentage points on the profit margin, and margins are getting skinnier and skinnier in the small fabrication world, so that has a drastic impact on whether you make the cut or not as an estimator.

Cliff234, see my comments further up about why I'm not going that route pre-bid.

 

[spieng89]

http://msc.aisc.org/globalassets/modern-steel/archives/2000/02/2000v02_rules_of_thumb.pdf

Good reference for structural rules of thumb

 

[KootK]

- You have a mechanical engineering background.
- You plan to run all of your designs by a structural engineer in the end.
- You're obviously pretty keen.

Given those things, I think that you should dive in and figure out everything that there is to know about designing steel beams. Then you can just do it. Perhaps a path like this:

1) Get a book like this: Link. It has a chapter on beams. You'll read it in a weekend or two and then know most of what there is to know.

2) Develop your own spreadsheet to tackle the designs. That will enhance your understanding and give you a tool that you know you're using properly.

3) For the first ____ beams, come back here and post your designs for review. We'll get you on the right path.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.