Computer Modeling - Level of Effort 1

[ash060]

Typically working on low-rise buildings (3-Story or less), I use hand calculations, spreadsheets, and Enercalc when performing calculations for design.

I have access to better analysis programs such as RISA and Ram Systems. I have noticed that these programs have the capabilities to design just about everything including masonry walls and steel joists.

I usually do not use them because it seems like using a backhoe to build a sand castle. I wanted to know at what point other engineers go from using hand calc's and spreadsheets to building a full blown computer model for their designs? Is there a big increase in efficiency to using the complete computer model for smaller building? Does anyone have a rule of thumb for switching from one analysis method to another (like a sq. ft threshold)?

Any input is welcome

 

[AELLC]

There is no rule of thumb - just user preference I suppose. For instance, I used to like Enercalc, but no longer use it...prefer spreadsheet.

When something is difficult such as too many load combinations, composite WF beams, etc, I would rather invest time in inputting a more intense computer program (past history when I was an employee working on a lot of heavy commercial stuff)

 

[ToadJones]

I guess it really depends on how proficient you are at using the modeling software.

I go to STAAD or RISA pretty quickly...I also like to use software and hand calcs/ spreadsheets together as a way to double check myself.

 

[271828]

When I designed buildings, I'd use a computer program to design just about any beam, column, frame, etc., even very simple ones, because so so many of them required redesign (usually LOL) due to elements outside my control.

If I design a beam manually, and the loads, span, etc. change, then the revised design is probably a complete re-work. Those changes are usually very fast with a computer program, such as moving a load 2 ft to the left, and running it again. Save the file over the old one, write the new size on the drawing, and move on.

The computerized revision is often a lot cleaner also. Using a uniformly loaded simple beam for example, say the span increased from 20 ft to 24 ft. Sure, I can scale up the moments by the ratio L^2 and the deflections by the ratio of L^4, mark out the old ones, and select a new beam pretty quickly, but that's going to be hard to follow months or years later.

So many parts ended up redesigned on my projects that it would've been much, much more maddening and slower with manual calcs.

 

[spats]

I agree with 271828... changes are a good part of the reason to use modeling software. RAM is so easy to use, why wouldn't you. You can set up a three story model in less than one hour. It chases all the loads down for you, including live load reductions.

One of the other big advantages in my mind is the ability to try out different senarios with little effort. Want to consider floor joists instead of beams? The change takes a matter of minutes, and the program does a tonnage take-off. Need to check floor vibration? Change the concrete thickness, concrete weight, and even beam spacing effortlessly. Want to design and do a take-off of WF vs. tube columns? Again, a matter of minutes.

And then when all is said and done, you have the confidence that you did not make a math error, and the program forces you to think through all the proper design parameters... sidesway amd k-factors, load combinations,bracing criteria,etc. And having complete and organized calculations all done for you doesn't hurt either.

As you can tell, I am a RAM fan. It's not as easy with STAAD or RISA as it is with RAM, but I wholeheartedly vote for modeling software. It's not cheap, but well worth the price.

 

[MiketheEngineer]

I'll jump to RISA for a quick truss analysis. Even though I can do those in my sleep - just makes feel more comfortable that I have done EVERYTHING right.


Of course - I am proficient with RISA and can load it up in matter of minutes - so it is worth the effort

 

[SteelPE]

I do work for a bunch of contractors... sigh

I will usually use RAM SS to do preliminary design. I don't really get to cute with the designs. I will use RAM to design the beams, columns and give me preliminary numbers on the LFRS. I will also use it to help with the design of joist with varying loads when I have multiple drift and RTU combinations but will not use it for simple joist designs.

I find it easy to run multiple building configurations for my clients. I am currently working on a structure that has had 4 different iterations in about 2 weeks. Each one gets sent to the contractor for an exact bid on the structural requirements. I am working on another project where the client wanted pile loads in 1 day on two different framing schemes (4 stories).

Once we are awarded the project, I will use RAM to back check the lateral forces in the building (I will run those by hand). I will then use STAAD to run a 2D frame analysis of the building (I don't like how RAM distributes the loads in the LFRS). If the building is a simple warehouse, I will use spreadsheets and tables to design the building and use RAM to check my analysis.

Pretty complicated I guess.

 

[pittguy12]

I also tend to go to the analysis program pretty quickly.

Sometimes I do go to hand calculate something. However, when it changes, it's just down right annoys me to have to redo them. Usually, I would have been better to invest the time up front to model it instead of having to change calculations and carry them through on my notes or spreadsheets.

 

[ash060]

Thanks for all the comments, but what about the cases for buildings that are not frame construction, but load bearing walls?

Is it easier to spend all that time working modifying wall panels with all the openings in the program to get the reinforcing necessary?

 

[SteelPE]

Well, the programs that I have do not have the load bearing wall design options, so I would just do those by hand.

 

[msquared48]

Anytime that matrices are involved in the analysis, I go to a computer program. 10 X 10's just take too long to solve by hand...

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[RFreund]

For those who model in computer program. Will use the program if the building is relativity small say 3 stories or so with masonry shearwalls? Do you model the diaphragm and connection is the shearwalls, ext.

EIT

 

[MiketheEngineer]

Might use the shear walls. Once I have the shear forces, I might design the coneections by hand??

 

[abusementpark]

I don't like how RAM distributes the loads in the LFRS

What don't you like about it?

I use RAM all the time for lateral load distribution. It works pretty well for simple diaphragm box structures.

 

[SteelPE]

What don't I like about it..... well when I have a structure that is embedded into the side of a hill the foundation walls tend to attract all of the lateral forces. I know that the program distributes the loads based upon stiffness and that this is probably an accurate measure of how much load will be driven through the wall..... I just don't like the way it goes about the distributions.

 

[abusementpark]

What don't I like about it..... well when I have a structure that is embedded into the side of a hill the foundation walls tend to attract all of the lateral forces.

Well geez, how often is that?? Maybe I live region with very flat terrain, but it's not too often I find myself dealing with a structure embedded into the side of a hill.

I know that the program distributes the loads based upon stiffness and that this is probably an accurate measure of how much load will be driven through the wall..... I just don't like the way it goes about the distributions.

So, I guess math and physics be damned?

Do you accurately model your diaphragms?

 

[RFreund]

When you say 'accurately model your diaphragms' what are you referring to (stiffness?), how do you guys typically model your diaphragms. Any input on RISA 3D specifically? I typically fall into the hand calc camp but should really start setting up a model more often. Although most of our work lately seems to be additions/remodel.

EIT

http://www.HowToEngineer.com

 

[slickdeals]

I think one of the biggest advantages of putting things in a model is that it is relatively easy to make the changes when the architect decides to change his mind (which is *too often* these days).

However, simple design problems are often extremely complicated when you throw it into a design program.

A FEA program will not do tributary area based force distribution, without fudging the stiffness modifiers.

 

[SteelPE]

abusementpark

Do I accurately model my diaphragms..... I try to do the best that I can. How do you accurately model a flexible diaphragm in RAM? I don't think RAM has that option..... or maybe I have been using it for so long they have now added it and I just haven't figured that out yet.

I also have many projects that are embedded into the sides of hills.

I love RAM.... I use it for gravity load distribution all the time especially when I have those pesky drift loadings. I take the frame distributions as an advisement to make sure my numbers are somewhat accurate. I just like a little more control during my frame analysis (which is why I use STAAD).

Don't worry, I never have any members that end up failing in the completed RAM model but passing in the STAAD models.

 

[spats]

RAM is capable of analyzing structures with rigid, flexible, and semi-rigid diapharagms. With rigid, of course, distribution of lateral loads is based on relative stiffness of the lateral load resisting elements. Flexible diaphragms are not capable of transmitting torsional moments, and distribution is based on tributary areas. Semi-rigid diaphragms are somewhere in between rigid and flexible. The program meshes the diaphragm, and values are input to model diaphragm stiffness.

In all cases, diaphragms can ever be sloped, and you can have multiple diaphragms of the three catagories at each story level.