How do you calculate the force in a compression flange? 2

[Ozy001]

How do you calculate the force in a compression flange based on the moment?

Hi,

I am trying to calculate how much force is in the compression flange of a portal column based on its moment. I have to design a torsional restraint (capable of withstanding 2.5% of the force of the compression flange). The angle of the tie from the compression flange to the restraining member is very accute and as such I am a little worried about the steel angles which have been added. Was thinking it would go something like this:

Calculate the Z vale of the compression flange only.
Divide the moment by the Z value to get a stress.
Multiply the cross sectional area by the stress. That should give me a maximum applied force.
Find the resultant force based on the angle of the tie.
Check the angle against the maximum force in compression to see if it passes.

Any help would be really appreciated.

Thanks!!!

 

[rowingengineer]

one vote for hokie's method here. if you try to use actual values for everything you design, you may as well give up now and become an accountant. Most if not everything in structural engineering as an approximation somewhere along the line, even if you don't know about it. for example the code rules for buckling of the flange brace.

How could you do anything so vicious? It was easy my dear, don't forget I spent two years as a building contractor. - Priscilla Presley & Ricardo Montalban

 

[Lion06]

I wanted to add to my above post. Using reasonable and valid approximations is a good thing, as long as you understand to do the problem correctly, understand exactly what you're approximating and why, and make the determination that it is, in fact, reasonable and valid to do so.

As rowing notes above, almost everything in structural engineering is approximated on some level.

 

[asixth]

Here's a quote you can take with you:

"If you understand why what you’re doing works, you’ll be much more effective at it!"

 

[ToadJones]

The OP, refreshingly, wasn't asking for the answer, but rather the correct way to calculate the force.
I tried to oblige.
However, If I was designing a brace like this (and I have many times) I would use the yield stress of the steel in the calc.
You'll find after designing a few such braces that 2.5% equates to a very small, easy to accommodate, load.

 

[WillisV]

You might also find that 2.5% might have nothing to do with what is actually required for bracing strength or just as important bracing stiffness - both of which should be calculated using AISC Appendix 6 for the particular situation in question.

 

[Ozy001]

As TJ has mentioned I wanted to find out how to do it correctly and not just guesstimate. There is a correct way to approach the problem and that is what is I felt important.

Lion, I fully appreciate that it would take you much less time if you just approximated but for the sake of 10 minutes I now have an accurate answer, a method for carrying out this type of analysis, a better grasp of what value of the applied load I would need to start worrying about, etc.

Being conservative is one thing but actually knowing how to carry out the process correctly is far more important for a graduate engineer don’t we all agree?

On that note I just want to post a thank you to TJ for his help.

 

[ToadJones]

you may wan to look into WillisV's comment as well.

 

[ToadJones]

....although from you OP, it appears as if you are not using AISC codes.

 

[SteelPE]

I also agree with the calculation provided by Hokie66.

If you have ever designed a moment connection, the force in the flange is M*12/(d-tf). This calculation is littered throughout the AISC manuals. So if it is not exact then why hasn't the AISC revised the calculation in all their design examples?

 

[ToadJones]

The OP is bracing a column, which will most likely also have axial load, ... axial compressive stress.

 

[graybeach]

I would definitely do Hokie's method. It there was also axial force in the column, I would divide that by the area of the whole section, multiply by the area of the flange, and add the result to M/D. This method is really accurate enough.

I would not necessarily use the yield stress times the area of the flange because slenderness might have controlled the allowable compression in the column.

 

[rb1957]

if you're designing the connection using yield stress*area is conservative, 'cause the load can't be bigger than that.

using yield*area as an allowable can be unconservative if the flange has a lower crippling/buckling allowable.

if you're just getting started in the business, i think it's a good idea to calc the load the long way a couple of times to learn how conservative is the yield stress simplification and how much (or little) time it saves, and to be carfeul in applying it.

it might be a "clever" way to show the calc and then to use the higher yield stress load as a conservatism.

the long method, as pointed out above, is the combination of axial compression and bending stresses.

 

[hokie66]

Ozy001,
Not much of what we do has to do with "actual values". For instance, the 2.5% used in your example is arbitrary, and could just as well be 3, 4, or 5. Some codes differ in the amount of bracing required.

 

[Ozy001]

But it has come from somewhere, I mean someone didn’t just wake up one day and decide that’s the figure to use. As with all elements in codes they have been based on sound engineering knowledge, much more specialist than anyone here is involved.

There is a correct method of carrying out a calculation. That’s important for a graduate, not just to quickly scoot around it and get a rough answer. I like facts, clear and in black and white. Even if the people responsible for writing the codes are completely wrong I am sure the insurers etc would be willing to accept that based on the code and through proper analysis my design was safe and proper.

Until I get to a stage where I am comfortable in my analysis I will always do the correct calculation to the button and not just approximate.

 

[asixth]

I think the technical paper it was derived from by published by trahair. I'll try and find and post it when I get home. Keep going though, it's good that your eager and trying to learn.

 

[graybeach]

Ozy - Can you post a sketch showing your column and the loading on it? It would be interesting to see how the various methods work out.

 

[nutte]

Ozy, don't write off the advice of those who seem to be simplifying the procedure. In some cases, there is a reason to use M/d [or more precisely M/(d-tf)] to calculate the flange force resulting from the moment. For moment connections where the flange connection is all that is being used to resist the moment, calculating the bending stress (M/S) and multiplying that stress by the flange area will result in a lower flange force than is actually there. Also remember that despite your desire for learning the true theoretical approach, there is a lot to be learned from the practical recommendations of the gray-haired.

 

[Ozy001]

Nutte and all others, I really hope I havent come across dismissive. I really appreciate everyone's input, I've already learned lots by listening to your points of view and from now will try to get involved within these forums as they are a great place to bounce ideas around! I've been at Uni for a few years but I have learned more in 6 months working with the my team (engineers and technicians). I know that the best people to learn from are experienced engineers. My original post was how do I calculate and not how do I approximate, as I thought it best that as a very junior engineer this is something which I should be able to do to the Newton. If however through time I get a feel for the values I'll never carry out the calculation in it's entirety. I just like to know that I'm doing it the right way and not the roughly way.

Graybeach, I'll upload the sketch tomorrow as I'm out the office now.

Once the sketch is uploaded and the variables and unknows set I would really appreciate the clear and definitive approach to the problem.

Once again all your help and advice no matter how little or expansive is always appreciated (that gos for the oldies too lol)

 

[handex]

Ozy, as people have said its great that you want to learn for yourself and understand whats going on. I fully advocate doing the proper and "Accurate" calc a few times so you have a feel for the values.

However, once you have been out of university for a while, you will being to realise that the best engineer is not the one being the most accurate or carrying figures to 6 decimal places. The main things is always understanding all the simplifications you make, and ensuring a safe and logical load path for all applied loads.

I did my thesis under one of the authors of the Australian wind code. I called him up one day to ask him for advice on a condition not covered in the code. He said "handex, your guess is as good as mine. Just make a conservative approximation".

Once you get on site and see how a lot of things are done, you will realise that the following quote holds a lot of truth for engineering calculations (quoted from a user of these forums)..

"We tend to measure with a micrometer, mark it with a crayon, and cut it with an axe"