More than one column on a baseplate. 2

[Leftwow]

Good Afternoon Gentleman,

Have you ever encountered a situation where you have 2 columns extremely close together, where you could combine baseplates? The first thing that comes to my mind is that the columns could twist the baseplate and cause some strange eccentricities. Have you done this before? If not, what are your thoughts?

 

[TehMightyEngineer]

If I recall the last time I looked at this the potential small material savings was greatly offset by the increased amount of complexity and potential errors in fabrication.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[Leftwow]

Yea man, the problem is, my Columns are Failing under the 2.0 Recommended strength factor K, So we are trying to make moment resisting base plates. Which brings my K value down to 1.2. There is four columns together, we need bolts on the outside. Honestly I don't think it can be done with the dimensional geometry available.

 

[TehMightyEngineer]

Oh wow, a 4 column baseplate expected to provide fixity? That doesn't sound very easy or cheap. Also, is sounds a lot like you're using approximate effective lengths. What's the project? Is this a building structure?

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[Leftwow]

Industrial Skids, kinda like a building but it has modular sections, also no cladding.

 

[Leftwow]

Mighty what is wrong with approximate effective length values?

 

[SAIL3]

one solution is to connect the cols to each other @ the base so they act as one composite col as far as the base pl is concerned....

 

[TehMightyEngineer]

Mighty what is wrong with approximate effective length values?

I thought you were designing a building and was about to chide you for not using the direct analysis method or otherwise considering P-Delta effects. Not as applicable to a non-building structure where second order effects are often negligible and your approximate effective length value is totally appropriate.

It's been a while since I considered multiple columns on a baseplate but I suspect SAIL has the best way to approach it. Support the columns on some sort of grade beam, dunnage, connection system thingy that makes them act as one unit and then provide a baseplate for that single unit. Will also have the benefit of reducing your unbraced length of the columns. However, I doubt it will be any cheaper than just up-sizing the columns.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[thaidavid40]

Multiple columns on a single base plate are common in heavy industrial buildings. It is common to have opposing lines of crane columns on either side of a central building column. However, these are usually aligned on a single common axis. If you are looking to combine four modular columns at a corner, that would create two lines of symmetry. I guess you could analyze the eccentricities in each direction, and then superpose them for their overall effect.
Dave

Thaidavid

 

[Leftwow]

Thank you for your responses.

Mighty,
I have never done a building more than one story, never had to use those methods. I would like to learn... and resources are still available. I will learn this eventually but I haven't had enough time and guidance as a young engineer to move on to those more complex methods. So I use the approximate values for my structures, which is what I learned in school a couple years ago. =)

Thai,
I am unsure how to go about doing that. I guess I would have to go back to the old structural analysis book.

 

[moltenmetal]

Any time I've seen four columns landing on a common baseplate, which I have ssen on industrial process skids which are multiple storeys high, the columns have also been joined to one another at levels above the baseplate. They are intended to act as much as possible as one column- always an interior column rather than a corner if you think about it.

 

[TehMightyEngineer]

Leftwow, here's a FANTASTIC lecture on effective length factors, some of the issues with it, and makes a good intro leading into the direct analysis method.

http://media.aisc.org/NASCC2014/140327-N40A.mp4

Here's a ton of examples for the direct analysis method:

http://media.aisc.org/NASCC2015/350325-N18A.mp4

There's a ton of great videos on this from other NASCC lectures, you can find them all here:

https://www.aisc.org/searchtaxonomy/conproceedings.aspx?id=4424

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[cal91]

I thought you were designing a building and was about to chide you for not using the direct analysis method or otherwise considering P-Delta effects.

I didn't know it was a bad idea to use effective lengths for column design for multiple story buildings! Guess I better watch the lecture also, thanks for posting!

 

[TehMightyEngineer]

Well, it's only a bad idea if you're not considering second order effects (either through amplified first order analysis or other methods) or not calculating your true effective length if you have sway frames. But, given how most (all?) multiple story buildings these days are designed on a computer, and most modern canned programs have the options to apply the direct analysis method, I fully think the direct analysis method makes the most sense and thus you can set K = 1.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[JasonMcKee71]

One interesting thing would be how to design such connections, as regulations don't cover these situations.

Jason McKee
proud R&D Manager of
Cross Section Analysis & Design
Software for the structural design of cross sections
Moment Curvature Analysis
Interaction Diagrams
Reifnorcement Design etc.

 

[cal91]

TME,

I've gone over the videos you showed and I've got a question:

Lets say you have a 40'-0" column you're analyzing with DAM and you split it into 4 segments. For strength checks of the column would you use KL=10'-0", or KL=40'-0"?

Is more information needed, such as bracing points, or do you not need to worry about bracing points since K is always 1?

 

[structSU10]

the DAM still takes into account unbraced length. If your 40' column has brace points @ 10', then that's ok, if its full height 40' then you should be using KL=40'. The plus for the method is not needing to determine K factors, which technically vary by load case.

 

[TehMightyEngineer]

Bracing is one of the most important considerations for compression. For a column, the L in KL/r is the distance between unbraced points; typically calculated for both the weak and strong axis (Lx and Ly) and compared to the radius of gyration for each axis (rx and ry). So, if you split a column into segments in a computer program you need to ensure that the computer program is using the proper unbraced length. For your example you would use KL = 40' unless you braced it at a distance less than 40'.

This is the same for both DAM and the effective length method.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

http://www.americanconcrete.com

 

[Sawbux]

I've done 4 tube steel columns on one base plate for architectural feature. Threaded glulam beams both ways at the top.

The columns would have to be tied together somewhere near the top, otherwise the plate will get bent in transport or erection.

 

[cal91]

Makes perfect sense, thanks.

One more question:

Even if you have a cantilever column (like a flag pole), you just apply the notional loads, reduce the stiffness, and still have K=1.0? I watched all the videos you supplied (thanks for those!) , but would just like to confirm this.