Is it important getting symbolic solution for structural? 4

[Siarhei]

Dear Colleagues
It would be very good to ask about one non usual task in structural engineering and its possible solution. I mean such problem conditions.
Please imagine structural construction from big number of rods, spatial beams and so on. It requires calculating it i.e. to find all internal factors, such as stresses and so on.
All today approaches (software, methods, and algorithms) offer only numeric solution for this task.
It will be very interesting to know has it interest for engineers and researches to calculate such structural system with answer in symbolically and numeric-symbolically notation.
For example, as we used always: for this type of loading 10000N we will have at this point of structure vertical deflection 0.9mm. Instead we will speak about result as expression for vertical deflection depend upon all system parameters like Segment Lengths, Young Modulus, Moment of Inertia, Section Area and so on.
Is it useful for researchers and engineers in real tasks?

Heartfelt Gratitude for any comment by this topic.
Eng. Professor Arlou

 

[corus]

I don't see how specifying a displacement in terms of the area of the section, for example, can be of any use as the two may not be related in a complex structure. Generally such results are based upon relative stiffness for which no single result can be plotted.

corus

 

[GregLocock]

I think what he's getting at is a complete analytical solution for a given structure, into which you could drop the section properties.

Yes, there would be some virtue in that if you were interested in optimising the structure. But if you are really interested in optimising structures you tend to end up changing the shape as well as the section sizes.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[IRstuff]

Frankly, I don't see any utility in something like that in general. Most engineering problems are too complex for symbolic solutions. As in Schroedinger's equation, anything more complex than a hydrogen atom does not have a closed form solution.

Additionally, sometimes, it's not just the actual solution that's difficult, but the structure of the problem statement itself, i.e., how would one even begin to create the symbolic solution for a 4-story building?

TTFN

FAQ731-376

 

[Siarhei]

Hi Colleagues,
Thanks for notices. Let's take a concrete sample for more subject speaking.
Of course, there is no 4-story building but it is real engineering task.
So, is there real value of such symbolic solution?
Deflection at center of structure.

 

[ThomasH]

Hi

I think that what you present as a "symbolic solution" in many cases whould be considered as a "formula" for calculating for example the deflection. In many cases the "formula" has a value for parametric studies. But what you need for the actual design is the deflection in meters or inches or whatever you prefer.

In general I would say that the actual deflection of the beam is of more interest than the equation giving the deflection. This might also be the reason why numerical methods are so popular. They basically give you what you need to do the design. Provided, of course, that all the boundary conditions are correct. But that is required for both methods.

Just some thoughts

Thomas

 

[ThomasH]

Hi again

Just an addition to my previous post.

As an engineer you need the results. How you got them is not neccessarily of interest provided they are correct. It's on the results you base you conclusion, "Thicker plate or not".

As a teacher you are probably not very interrested in the students results in terms of x mm deflection. It's how he got them that is of interest to check his work.

Now, when I work with junior engineers I'm interessted in both the results and how he/she got them. This is to verify that they are correct. But typically it would be a numerical solution. A symbolic solution is often nearly impossible to get within a resonable timeframe.

Some more thoughts

Thomas

 

[civilperson]

Truthfully, the deflection example is poor. No one cares about total deflection of a series of beams as long as the individual defections of the components meet some arbitrary criteria, (say L/600 or L/360). Loads are approximations and only live load deflections are of any interest. If the the loading is maximized everywhere, then it does not approximate the real world. This is a academic exercise for the student who aspires to teach!