Compound Sections 1

[LSPSCAT]

I am looking for information on analyzing compound sections that are not connected; simply stacked. Examples are stacked plates for load distribution or vertically stacked wide flange sections in bending. It is in the area of construction and these items are frequently used based on availability.

 

[corus]

Well a lot of flak here!

In fact I'm a mathematician and not a mechanical engineer and so tend to think a tad mroe about problems than perhaps the simple engineer with his simple text book.
All of you are wrong of course because you forgot to consider friction between the two sections. In a perfect frictionless world then you'd be right and the two sections would slide across each other and you'd consider the two sections separately. In the real world, folks, there is friction and as such the true answer will lie between the (impossible) case of frictionless surfaces and (the more likely) case of rough surfaces when sliding will be prohibited and the two sections will act as one. A finite element model, that included contact between sliding surfaces, proved that theory for me. If anyone can do the same caluclation by hand then I'll throw awyay my pc.

As I said previously, the worst case would be to consider frictionless surfaces so that would be used in any design calculation. As engineers though you'd probably multiply the answer by two, just to be safe.



corus

 

[civilperson]

Corus, Which type of friction does your model use? After the initial movement between the two sections, does any residual friction exist when both have deflected?

 

[nutte]

Corus, from a theoretical standpoint, you're right. There's friction, so there will be some transfer between the two members.

From a practical standpoint, an engineer would assume no friction and distribute the forces accordingly.

That is the difference between scientists and engineers. The more exact solution in this case would likely yield negligible benefits. And when you add your factor of safety to it, which you seem to have a problem with, the benefit is reduced even more.

 

[miecz]

corus-

please, don't give our code writers any ideas...

 

[Lion06]

I am not so sure that once it begins to slide that friction isn't overcome completely (or pretty close to it). Either way, it is certainly not anything that can be counted on with ANY reasonable degree of accuracy.

Corus-
Mathematician or not, the idea that this surprised you gives you no room to talk about us "simple engineers". Additionally, if you have so much mental superiority to us, why are you on this forum?
Go start your own math forum.

 

[DWHA]

Corus-
I like how you say: "the simple engineer with his simple text book" because in my opinoin, mathematicians are a lot more text book oriented than engineers. We have to make things work in the real world. Everything that we do on paper, in the end goes out and is built in the real world.

 

[Lion06]

corus-
Additionally, even in the case of the roughest surfaces, you would need a coefficient of friction that is completely unachievable to get full composite action.

 

[civeng80]

Unless Im missing something here, if the sections are not connected there is no horizontal shear transfer so that the sections cannot be designed as compound sections. Each section would take an equal propertion of the load.

 

[Lion06]

civeng80-
I agree, but corus insisted on trying to make a point.

 

[corus]

StructuralEIT,
I never said they would act as a full composite section, nor do I say they act separately, but in the real world they would act somewhere inbetween due to the friction. If you're looking at structures and their behaviour then it's important to take into account all the forces. This is quite often not the case as the simple solution is to ignore these complications and just do a simple hand calculation. Unfortunately ignoring such factors can have implications that can lead to failure elsewhere, and that's why using even a simple FE model can help your understanding of the fundamental concepts. Don't knock it.

corus

 

[Lion06]

corus-
I don't have the time or energy to argue with you about this, but you will not find an engineer anywhere that would count on friction causing composite action in a case where the sections are mechanically connected except for a vibration check and possibly a couple other unique circumstances. It would certainly not ever happen in a strength limit check.

 

[Lion06]

Corus-
On second thought, just for the heck of it....... please answer the following questions very specifically, please do not dance around them with vague, general answers.

1. What coefficient of (static) friction did you use?

2. Where did you arrive at this value?

3. Once the members began to slide, how much residual friction remained once the sliding stopped?

4. Did this take into account the level of acceleration of slip between the members once the members started to slide and the amount of slip due to it?

5. What coefficient of (sliding) friction did you use for that calculation?

6. Where did you arrive at this value?

7. After all of your convoluted assumptions (your assumptions of the coefficients are just as much of an assumption as us taking it as zero - whether you prefer to admit it to yourself or not. After all, there is no way you can no with any certainty what they (without testing, which you can't do for every member) are and if you choose to be conservative then what the heck is the point of not using zero?), did your FE model arrive at an answer that was more than about 2% - 3% different than neglecting friction completely?

Can you please give me an example of how ignoring friction between "partially composite" (your term, not mine) members can cause failure elsewhere?
Also, I certainly don't need an FE model to help my understanding of how friction affects the interaction of two beams that are not otherwise connected. If you do, then you have issues with the fundamental concepts, not me.

 

[Lion06]

corus-
One more thing.... how did you account for the varying normal force (and resulting varying frictional force) along the length of the beam? Certainly a simply supported beam (2 stacked on top of each other) with a point load at midspan would not have a uniform normal force between them.
If you assume a uniform loading (such that you may be able to count on a uniform frictional force), then you have a varying shear flow, not the max shear flow that we assume at the end of the beam.
There are a million and one things that CAN be taken into account. That doesn't mean that all of them SHOULD be. Engineering allows us to make slightly conservative assumptions in the interest of time when the "true" solution would take a ridiculous amount of time and still employ A LOT of assumptions. If you are going to be making assumptions, why not err on the side of caution?
I doubt someone who was on a floor that collapsed would appreciate the excuse, "Well..... we counted on some friction that may not have been there to save the owner $100 in steel weight".
Come through the door that little door on the right labeled, "to the real world".

 

[IFRs]

This debate is not so much if friction exists but rather if one should take friction into account when designing these type of compound sections. Typically, engineers make simplifying assumptions. Ignoring friction in this case would be one. The fact that it is ignored in the analysis does not mean it is not present, just that it would not be cost effective to include it in the analysis. We like to make simplifying assumptions when the result will be conservative, and in this case it would be as any friction would tend to make the structure stronger.

 

[DWHA]

Arguing with an engineer is like wrestling a pig in the mud, after a while, one will figure out that the pig enjoys it.

With that being said. I think that we have to come to a generally agree and to include friction for the design of a compound section is not a common practice, and should not be included in 99.9% of cases. I hope that we have answered the question for jchc and this may be a good time to end this discussion.

Oh ya, Structural EIT, good quote “Mathematician or not, the idea that this surprised you gives you no room to talk about us "simple engineers". Additionally, if you have so much mental superiority to us, why are you on this forum?
Go start your own math forum.”

 

[JAE]

OK, let's all calm down here a bit eh? This is Eng-Tips, not some political blog commentary page.

corus, you said:

For your worst case I'd assume you have frictionless surfaces and apply half the load to one beam.

Which is what StructuralEIT is advocating. I agree with this statement and for practical structural design, you simply don't count on the friction because StructuralEIT is correct that you cannot legitimately account for it quantatively.

[blue]From theory - yes, it is possible to consider friction.[/blue]

[red]From pragmatic practice - dumb to do so because if you did your hand would be shaking so much when you went to sign and seal it that you couldn't sign - and you'd risk peeing your pants.[/red]

 

[40818]

Calm down certain people.
Its christmas eve, and i'm sure everybody has better and more enjoyable things to do than argue in this manner.
Debate is one thing, but we all need to keep it polite and respectfull.
For my pennys worth, structuralEIT is correct. But Corus is also correct under his own assumptions, the arguing is about the validity of the assumptions.
I dont know Corus, but his answers have always been decent, and he knows his FE stuff, he has answered 383 in Abaqus and 351 posts in FE. So its an insult to him to even question his technical FE ability, especially from somebody who has never posted a reply in a FE forum
Were all professionals here so lets keep it that way.

 

[miecz]

corus

When entering a thread, perhaps an opening other than:

In fact I'm a mathematician and not a mechanical engineer and so tend to think a tad mroe about problems than perhaps the simple engineer with his simple text book.
All of you are wrong of course...

Such a post can produce friction where there otherwise isn't any.

 

[civeng80]

One further comment for the record !

Ive got quite a few textbooks on solid mechanics and strength of materials including the work of the great Stephen Timoshenko, as I'm sure my esteemed colleagues have in their libraries. In my opinion they are not simple textbooks, but they are practical and given the (reasonable) assumptions made do give quite resonable answers. Mathematically more precise models of course may be built e.g. by our colleague Corus, I for one as an engineer have complete faith in these textbooks and I still proudly refer to them from my bookshelf.

A Merry Christmas to all !