high moments at column/base 1

[oneintheeye]

hi, when modelling column loads on a foundation slab and the column loads include a moment the model will show high moments local to this applied load. How do you cater for this? I have had a few suggestions, some seem resonable, some i've used in the past but would like to hear the experts view (you guys).

thanks.

 

[oneintheeye]

refining the meash doesn't seem to help much. The effect is obviously greatest if the moment is applied as a point load. The other option as someone suggested is use tension or downforce in the bolts. But as bolt spacing will be reletively small even then when we have high moments the plate still 'twists' considerably.

 

[patswfc]

Herewegothen,

I've encountered this problem many times with FEA analysis also. When you apply a moment as a point load to a slab you will get very local forces in the slab. Spreading this moment into a couple in anchor bolts will help a bit but not much, as the bolts wont have a big lever arm. Refining the mesh will only make things worse.

Normally the FEA program will give a contour plot of bending moments in kNm/m width of slab, with very high peak moments at the column. These will only act over a very small distance and their magnitude will drop of very quickly as you move away from the column.

What i generally do in such a circumstance is determine (using engineering judgement) what width of slab is effective in resisting the moment that is transferred from the column. Then average the moment over this width and design the slab/beam for the averaged bending moment.

Hope this helps and also hope that this is how you have dealt with it in the past.

 

[oneintheeye]

yes I have done something similar when dealing with very local moments. Sometimes we have very high moments associated with the columns which makes me uneasy doing this at times.

 

[miecz]

I've experienced this as well, and have averaged the local moments over an effective width, based on engineering judgement. I was hoping you would get a response with a reference to some research. Seeing none has been posted, looks like we're on our own with this kind of thing. I believe I have used an effective width of 3 times the thickness of the concrete to each side of the applied force. I wonder what others have used.

 

[oneintheeye]

ive resticted mine to average over a metre really. though I suppose you could go further, effective depth either side of column? so 2 X eff. depth + column width? Looks like its a, as you say 'engineering judgement' call.

 

[fasboater]

herewegothen,

I don’t quite understand, you stated a couple of threads ago that you have "pinned supports". How would transfer any moment from the proposed framing (column) into the foundations?

If you do in fact have a fixed base column that is transferring moment then you could go use Steel Structures by Salmon and Johnson. Using the Steel Structures book you would essentially convert the moment into a moment couple (I know it’s the foundation your interested in, not the steel). By designing using this procedure (grouting under the base plate) you basically reduce your loads into a bearing load at the top of the foundation and a tension load through the anchor bolts. You can then go to your ACI code and check the design (anchor bolts for pullout, the footing punching shear etc.)

Typically based on the anchor bolt size, length and as you mentioned moment arm you can transfer a fairly large moment. The only issues that I have run into when your bolts are fairly close together is that the pullout cone overlaps and the concrete does not yield a linear correlation between number of anchor bolts and pullout strength. You take a pretty good hit with groups of anchors.

I hope this helps.

fas

 

[oneintheeye]

pinned or spring supports were referring to piles position not column from superstructure. The approach you state I have used in the past. The fact that bolts are always reletively close together ends up giving the same effect in the top of the slab as just applying the moment. If you run through a FEA program and apply moments you will see the problem.