Properties of Combination Sections 8

[vintage70]

Aside from the Tables in the AISC Manual
where can I find more variations of beam and angle combinations? Is there a way to calculate S1 & S2 & I?
Are there any programs available?

 

[whymrg]

In RISA-2D or RISA-3D you may specify built-up section and it will calculate properties.

Or you may use RISA-Section (I have not used it, but my coworker had and said that it was good).

 

[VoyageofDiscovery]

Try ShapeBuilder 3.0, but keep in mind, that if the in-service section is loaded there is an initial locked in stress after the section is built up.

 

[whyun]

Shapebuilder is a good program. Once the built-up section is created, it can be imported to the VisualAnalysis program for analysis and design.

For a stand-alone program, you may look into one of the modules within EnerCalc.

 

[LPPE]

In AutoCAD, draw your cross section. Convert it to a REGION. Then do MASSPROP, find the centroid of your shape, then move that to 0,0,0. Then do MASSPROP again, and that will give you some of the values you are looking for.

 

[austim]

in 45+ years of practice I have never needed to use software for such a calculation. If your section will be symmetrical about the plane of the web, then what you want to do should not take more than 5 minutes of hand calculation.

Why not just go back to basics and find a textbook (or your university notes ?) which shows how to combine section properties?

 

[lippie]

I agree with austim on simple shapes. STAAD makes a product called Sectionwizard for complex shapes. I hear SAP makes one as well. But since STAAD is the industry standard, it probably would make sense to go with them. More compatibility.

 

[austim]

Hi lippie,
My post was aimed specifically at the very simple problem that vintage70 posted.

Essentially this only requires the combination of the section properties of two basic sections (The main I beam and the pair of angles), whose individual properties are known from standard section property tables, and which have a common axis of symmetry (so we know from the start that the principal axes are horizontal and vertical).

I cannot lay my hands on a text that sets out the method of calculation, so what follows is as good as I can offer. Please excuse the limitations that are inherent in these posts (I cannot, for example, set out the neat tabulation of values that I was taught to use back in the 1950's).

Let me start by defining my variables:

Datum line : I usually use a horizontal line through the bottom of the I beam as a datum, from which all vertical offsets are taken. (Some prefer to use the NA of the main beam as the start datum - there are arguments in favour of either choice).

Variables:
Acomb = total area of combined section; Ybar = distance of NA of combined section above the assumed datum line;
AYcomb = 1st moment of area of the combined section about the assumed datum;
Icomb00 = "Moment of Inertia" (or 2nd moment of area if that is what you were taught to call it) I of the combined section about the datum line;
Icombna = I of the combined section about its own Neutral Axis;
ybtm = distance from combined NA to bottom fibre; ytop = distance from combined NA to top fibre.

A1=main beam area; y1=vertical distance to centroid of main beam; I1 = I of main beam about its own Neutral Axis;

A2=area of pair of angles; y2=vertical distance to centroid of angles; I2 = I of pair of angles about their Neutral Axis;

For easy legibility of any formal calculations I tabulate 6 columns for the following properties for every component section involved - Description of component section (with dimensions where appropriate),Area, y, Area*y, Area*y^2, I.

Particularly if you are calculating somewhat more complex sections (eg a PS concrete I beam, with triangular fillets between the web and both flanges) the tabulation keeps everything clear for anyone else looking at the calcs.

Basic equations:
Acomb = A1 + A2 = sum of my table column 2
AYcomb=A1*Y1+A2*Y2 = sum of table column 4
Icomb00 = A1*Y1^2 + I1 + A2*Y2^2 + I2 =sum of table columns 5 and 6
[for the more complex sections, these just become Acomb=A1+A2+A3+A4+..etc]
That is the hard work done .

Now we get ybar from ybar=Aycomb/Acomb.
Icombna=Icomb00-Acomb*ybar^2

Here comes the benefit of taking the datum through the bottom of the main beam:
ybtm = ybar, Section modulus to bottom = Icombna/ybtm.
[with any other datum you have to calculate the bottom fibre distance as the distance from the datum to the bottom fibre of the main beam plus ybar]
ytop = Top fibre distance = overall depth - ybtm, Section modulus to top = Icombna/ytop.

All done. In fact it takes much longer to describe than to do on paper. Total calculations involved: 2 mental doublings (single angle area and I to give properties of the pair), 4 multiplications (Ay and Ay^2 ), 4 column additions (A,Ay,Ay^2,I) plus 4 simple expressions.

 

[trilinga]

austim has explained the process in detail.
The entire exercise has only two steps, viz. Finding the CG of the combined section and find the MI of the combined section about the axis passing through CG using the formula

Icg = sum(Icg of individual elemnts+ area of element * d^2) where d is the distance between the CG axis of the element and the axis through the CG of the combined section.
As austim has stated this geometrical calculation does not take much time to do by hand.

I am afraid that the computer era is eroding our patience and confidence to do simple hand computations!

 

[AJUK]

I think there is a section property calculator (with some other good free stuff) on this link

http://www.structural-engineering.fsnet.co.uk/free.htm

However, I do agree that we should not place too much reliance on computers for such a simple calculations.

 

[vintage70]

Thanks to all. Especially to you austim. I usually encounter simple beams or symetric combinations. This one was a MC12 with a 10x1/4-in plate and a 2x2x1/4-in angle. So I was curious. Thanks for the explanation. I'll try it long way.

 

[trainguy]

I like Pylko's AutoCAD method. In 80% of cases, the shape, or a similar shape, has been drawn by somebody else, so there it is. 3 minutes later, you have all the props you're looking for. Be VERY careful, however, about checking AutoCAD's nomenclature and datums, etc. Just do 2 or 3 examples by hand, and then with Acad, maybe even combined with a very small Excel spreadsheet for repeat calcs, and I swear, you'll NEVER do them by hand again.

As a matter of fact, you can set up your spreadsheet to calculate your centroidal moments of inertia using the parallel axis theorem, and voila, a single step.

Then, when you find out your stress is still too high, or something else needs nodification, 3 minutes later...finished. Can the hand calc. proponents match that?


You can even document this nicely by incorporating the Acad view of your section an an Excel document.

 

[austim]

I have always done that type of calculation without needing access to a power point (occasionally in-flight and with a slide rule)and with total equipment and software cost less than $10. Can the computer proponents match that?

Must we really assume that all engineering design has to be done at a desk with full computer and software support and that it is quite OK for engineers to forget the basics?

 

[jheidt2543]

When I was a young engineer, I used an abacus, both ways, up hill, in the snow! <G>

Really, just a little poke. I agree, the computer is great, but sometimes just setting the problem up takes longer than solving it by hand.

 

[trilinga]

I agree with you austim.

The same question has been discussed in different threads under different titles. Should the use of computers and software blunt the mental working capability of an engineer?

 

[flamby]

1. Trainguy's spreadsheet is nothing but a Hand-Calculation computerised. You got to feed all the equations yourself, so techically equivalent to a hand computation.
2. An engineer forgetting the basics is not an engineer. If one becomes useless in situation where there is no access to computers, God shower His mercy on him. (And still there are many such places!)
3. If one 'Knows' how to compute sectional-properties, whether he chooses to use calculator, computer or a sheet of paper is really a trivial issue, as far as he gets the correct values.
4. I regularly use computer programs and also do hand computations. My limited experiencxe tells me, mistakes done on computer are most hard to find, even to experienced eys. Contrary to this a hand calc can be checked by a Grad engineer and he/she may be able to pick up any mistakes there. If I were to bet my hard earned money on some computation, I will most definitely do a hand calc. The satisfaction of doing every mathematical step yourself defeats all self-doubts and fears.
Regards to all,
flame

 

[trainguy]

When my projects finish up on time and on budget, because I prepared, tested and documented my spreadsheets for frequent re-use, I don't hear my clients touting the virtue of slide-rules. Of course I would never blindly rely on some purchased software package (complete with disclaimer)without doing a multitude of tests and examples.

I also understand how to calculate all the section properties that I need by hand, but why waste time over and over when the client is paying a fixed price for my services? It leaves me (and my colleagues) more time for checking and correcting drawings.

Being an engineer in this day and age requires BOTH an understanding of classical methods, and an openness to new technologies to speed up the process.

And this way, when I fly, I don't need to calculate section properties, I can just enjoy the movie and pretzels...

 

[trainguy]

Will the real Austim please step forward:

Are you the same Austim that suggested a non-linear analysis to analyse tension-only braced frames?

But you calculate section props by hand?

Hmmm...

 

[whyun]

i like the post by flame.

computer is supposed to help us do things faster. there is a tendency of many younger generation engineers to rely fully on computer out-put. cliche'd garbage-in-garbage-out remains for ANYONE who use computer programs. There had been many good articles I've read in the past about structural engineering, computer-use and engineering judgment.

True engineer can solve problems even after you take away the codes, computer and calculator. That is the future generation of engineers I would like to see.

I see many spreadsheets being passed around on the net. I never use any - not because I dont trust the author but I wouldnt have confidence in the spreadsheet unless I took time to know what is going on in EVERY CELL.

Even in-house, I am very careful about distributing spreadsheet tools which I've authored. Peers may not know the assumptions, limitations in the routine.

Bottom line is... Computer is merely a tool. Use it wisely, otherwise it will come back and bite you.

few cents from sunny California.

 

[JAE]

A few more cents from rainy Nebraska (not a husker fan, though).

I would like to boil down two concepts - one from austim and one from trainguy.

1. I agree with austim that engineers should know how to do the section calculations by hand. They should know the concepts behind them, and the process to find a section property without a desk/computer. However, in this current economic/engineering culture these days, speed is part and parcel of our services...not just our engineering expertise. As an engineer, I do not have the luxury of doing all my calcs by hand. I must use the technological tools available to me to provide my services to my client.

I'm not advocating engineers becoming blindly dependent upon the black box. I agree that a good engieer should use the CORRECT tool for the job...that may mean a computer program, spreadsheet, hand calc or counting on fingers...whatever will get me the answer, most correctly, in the shortest time.

All this while keeping myself cognizant of quality checking and assurance that I'm doing the right thing.

2. trainguy hits this same point - be open to the tools out there as they become available. BUT, don't use a tool that you know nothing about and have not tested.