ECCENTRICITY by cantilever?

[bdruehl]

question: should an eccentricity not be caused by a cantilevered beam on its supported column? seems to me that it should be, but i dont know/remember how to get at it (based on the worst case unbalanced loading). it makes a big difference in the column calcs.

 

[hokie66]

A cantilever causes bending moment. Eccentricity causes bending moment. The bending moment has to be designed for, in conjunction with the axial load.

 

[bdruehl]

thanks for comment.... i would agree with your statement.

well, ill make the question simpler... does a cantilevered load supported by a column necessitate the input of an eccentricity in say STruCalc column design?

 

[hokie66]

I have no knowledge of input conditions for that program.

 

[apsix]

If you model the column to beam connection as a moment connection you will be conservative (or accurate if it is a moment connection).
If it's modelled as a pin on the neutral axis you will be non-conservative.

 

[bdruehl]

wood support column to steel cantilever beam.. so no moment connection will be modeled... so its a pin type roller at the far interior support.

i guess i am not making myself clear.. i apologize... this is not a steel connection, it seams to me that a large cantilever especially if unbalanced loads are applied (which they should be), IS THERE AN eccentricity to the column that supports it (it is bearing on the top of beam) BECAUSE it is a cantivelever?, and therefore the actual vertical reaction to the column would be ofcourse skewed to cantilever side of the column, ie.. an eccentricty that i need to account for? that is the question. thanks! it seams there should be an eccentricity to me, but perhaps its ignored or i am in left field.

thanx!

 

[hokie66]

You seem to be confusing yourself. The cantilever load is indeed eccentric to the column. That is the meaning of cantilever. The distance from the centroid of the load on the cantilever to the centroid of the column is the eccentricity. But when you say "pin type roller at the far interior support", it leads me to think that you don't have a cantilever. Perhaps you need to provide a sketch.

 

[apsix]

I'm assuming your cantilever is a cantilevered beam with a backspan.

There is an eccentricity for ANY beam bearing on the top of a column, due to the rotation of the beam at the support.

The eccentricity is not necessarily more for a cantilever, it depends on geometry, stiffness and loading. Indeed, rotation at the support can be contrived to be zero for a cantilever, therefore eccentricity would be minimal.

My steel code, AS4100, requires the eccentricity for a simple bearing connection on the column to be depth/2.

My point about using a moment connection is that it would be conservative, therefore safe.

 

[BAretired]

If you want a sensible answer to your question, provide a sketch of the cantilever and the loading. I don't know about anyone else, but I don't understand the question.

BA

 

[hokie66]

Right. Sketches are good, or failing that, adequate verbiage. I took the question to mean column and cantilever, but now that apsix has chimed in, it is probably just a simple span beam with overhang. In that event, I agree with apsix.

 

[vandede427]

if the cantilever has a backspan, then i would assume no moment transfered into the column.

 

[bdruehl]

again, i appreciate the responses! and take my lashings for inadequate verbage and no sketch.

ok, this is a large cantilever (see my other post), 16 foot interior span and 16 foot exterior cantilver span where i am scabbing steel to fix rotten original glulams, one of which sits on an original WOOD post in wall. Loading is 1250LL, 125 DL plf on cantilever.

since the new code requires 100 psf live load for balconies, it is pushing the limits of the whole system. seems i incorrectly conjured up that when the cantilever is loaded to its max it would create an eccentricity on the post towards the cantilever side, a thought that i usually ignore and/or would never come up, but like i said this system is approaching max capacity so i dont want to "underthink" it..

i have a wood post so it wouldnt apply, but an eccentricity of d/2 definetely blows out the column. seems that eccentricity for top bearing is commonly ignored - as i usually do.

thanks apsix et al!
....

 

[BAretired]

So Span A-B is 16' and Cantilever B-C is 16'.

When live load extends from A to C, the post at B has maximum load but the beam has a very small rotation. The post at B must be designed for a maximum load of 32(1250 + 125) = 44,000# at minimal eccentricity.

When live load extends only from B to C, the beam slopes at Post B, but the total load is reduced. It would be prudent to consider a greater eccentricity on Post B under this condition.

Post A under unbalanced live load would have an uplift of 10,000# and needs to be held down to the foundation which, in turn, must be capable of resisting uplift.

BA

 

[bdruehl]

yes baretired.. that was my question and you are correct in your span descriptions and loading .. except that the live load on interior of house (span a-b) is 40 lb live not 100 - but as you say later the critical load is cantilever only loading.. ie a huge dance party on the deck.

.. the question is how much of an eccentricity and based on what analogy? it is critical because the post is existing, works with no eccentricity but fails with even a 1 inch eccentricity if i trust strucalc... usually i would "overdesign" the post so that any concievable e... unable to find any code related information on "required" eccentricity for this particular case.

 

[bdruehl]

this whole thing is related to this post, which is about all of the other issues of this cantilever design which i overthink...

http://www.eng-tips.com/viewthread.cfm?qid=266060&page=1

 

[paddingtongreen]

I'm not sure I understand this properly.

But here goes, anyway, I assume that, so far, the post is considered pinned at the base, and at the top. You say the fire code requires it to be wrapped. Would it help to make that top connection a moment conn., it would be hidden by the wrapping. This comes close to BA's solution but gives you the chance to use a stiffer beam to control the moment, bearing in mind that the modulus of elasticity of the post is much lower than that of steel.

I can't help feeling that I am talking out of the back of my head here, but I can't see what's wrong with the thought.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[BAretired]

You could calculate the rotation in the beam with hinged supports, then assume the same rotation in the top of the post. From that, you could calculate the moment in the post, hence the eccentricity.

I do not feel comfortable with a post which fails with a 1" eccentricity. What size is the post? Is there enough clearance to add material to the post?

You could replace the existing wood post with a steel post of the same size.

If there is enough room, you could leave the existing post in place and add a steel channel on opposite sides, bolting through the post at, say 2'-0" centers.

Or you could build up the post with timber planks on opposite sides.

BA

 

[bdruehl]

yes all of those solutions would be on the table..

i agree with your view of "if a post fails with 1" then its questionable. wish i could get a concrete answer on IF there is an eccentricity, and if so, what is the correct formula? calculate minimum bearing area required by load and skew the load from center to the center of the bearing area? nah, doubt it, would have to do with the actual deflection of the beam ... i think most heavily loaded columns will fail with any nominal eccentricity... but i dont know, eccentricty isnt always on my plate to think about. thinkin aloud again!

and padding, perhaps, but only teh beam itself needs to be wrapped, the post is in the wall... and i have had only very painful success in trying to do moment connections of high magnitude with wood.

 

[BAretired]

I don't think the question is IF there is an eccentricity. There is. The question is how much is the eccentricity.

If the beam is very stiff relative to the post, you will not be too far off if you calculate the rotation of the beam with unbalanced load, then apply that rotation to the top of the post. In other words, the top of post simply rotates to match the beam. The calculation is easy to make and it will give you a slightly conservative estimate of the moment in the post.

BA

 

[dhengr]

bdruehl:
You better not have a roller for a reaction at the back span, or you have an unstable col. in the direction of the length of the beam. That reaction better be tied down for uplift and prevented from moving in the direction of the length of the beam. I think I understand your concern though, you think that as the canti. flexes (its curvature), you will get an “e” approx. equal to (col. dim./2) toward the tip of the canti. as the outer edge of the col. is loaded by the bending beam.

Is this a public balcony with a LL of 100psf, or a residential deck with 40psf +, and a couple feet, right out at the railing, at 60psf? How do you know the size and stress grade of the col.? What is the load path at its base, a 44kip col. load is fairly sizable. I would reconsider a col. which couldn’t take a 1" eccentricity.

BAretired has it right in his posts 2MAR 15:44 & 19:08. As an approx. I would use a triangular loading to rep. the beam curvature. Thus, the resultant might fall at (d/2)(2/3) = d/3 from the center of the col. However, I think the bearing triangle would start inboard of the col. center so I’ll say maybe e = d/4. This would induce an eccentricity, and moment at the top of the col. but not much curvature at that point in the col., since I think you would just get some crushing of the wood parallel to the grain.

Why don’t you nail a 1" thick plate to the top of the col., same size a col., and watch nailing edge dist. Then weld a 3/4" square bar to the top/center of the pl./col., now e = 3/8", approx. zero. Now check web crippling on the stl. bm. And, I still don’t like a col. designed that close to its limit when I don’t know its condition or stress grade. You say it worked up till now, so it must be O.K., and I suggest it probably wasn’t originally designed for 100psf and just hasn’t seen its design load plus.