Inclined Column Question 1

[Ashke15]

Hi everyone, I'm looking at a situation I haven't encountered before and wanted to get some feedback if you have some time.

I've got a large custom home in a high snow load area with a covered deck at the back. The deck is 25' wide and has an inclined steel column at each end. There will be a beam at the top to carry the roof and a beam at the deck level. I have resolved the loads into axial compression in the column, vertical reaction at the bottom, and horizontal reaction at top and bottom using the D + .75S + .75L load combination. My question concerns the horizontal reaction at the top of the column. What prevents the beam from rotating as it tries to resist the horizontal force from the column?

The beam of course welds to the top of the column. It appears to me that the column tries to rotate the beam away from the house. The welds hold the beam in place, but then the beam tries to pull the rafters away from the house. Is this the correct load path? I need to make sure that the rafters are secured appropriately to the beam and back at the house so that the beam doesn't pull away from the rafters, and so that the rafters don't pull away from the house. That's how I see it, but wanted to see what you think. Do the deck joists see any of that force pulling them away from the house? Would appreciate any feedback. In summary:

1) The beam rotation is resisted by the welds from the beam to the column? I will also have web stiffeners.
2) The beam imparts a horizontal force on the rafters pulling them away from the house?
3) Do the deck joists also pull away from the house?
4) Recommendations from your experience?
5) Is there a certain degree of incline where you need to design the column with a moment at the base instead of only vertical and horizontal reactions?
6) Where did I mess up?

Thanks!

 

[BAretired]

I started to draw the structure, then realized that there should be a sloping column at both ends of the deck. Draw the whole structure, not just the right hand side. I don't understand what it looks like.

 

[JungleJoe]

Draw the whole structure, not just the right hand side.

OP said it's a 25' wide deck with an angled column at each end

 

[Ashke15]

Draw the whole structure, not just the right hand side. I don't understand what it looks like.

Thanks for your response. Sorry, I thought I could get away by only drawing one side of it and explaining that it's symmetrical on the other side. The drawing you posted is exactly what it looks like. The ABC line is the back wall of the home. The BE line are the deck joists. The CD line are the roof rafters. If you copy your drawing and move it 25' into or out of the screen then that's what the full structure looks like. The beam at the top (going into the screen at point D) and the beam at the deck level (going into the screen at point E) are both 25' long.

 

[SWComposites]

If it was me, I would add some bolted gusset fittings between the web of the beam at D and the rafters. Bolt thru the beam web to attach back to back fittings on the the inside and outside of the beam.

 

[human909]

6) Where did I mess up?

Nowhere so far. You are part way through the design and asking good questions your own design and now others. So I would say you haven't messed up.

-Just make sure you have clear load paths for the horizontal load produced by the column.
-In this case I likely wouldn't just rely on a welded beam with a few stiffeners.
-I'd have a short fly brace out left or right to a rafter and explicitly ensure that rafter can handle the axial load, design heavier rafters about the columns if required.
-Ensure your lateral system and roof diaphragm can handle the significantly higher permanent lateral loads being placed on it. This might involve specifically chasing the load path rather than the numerous simplifying code allowances regarding lateral loads in residential construction. (The code I use have them. Eg rigid diagrams and only designing the lateral system for wind.)

EDIT due to timing of posts:

Thanks for your response. Sorry, I thought I could get away by only drawing one side of it and explaining that it's symmetrical on the other side.

Well if it is symmetrical on the other side then you can mostly get away without beefing up your lateral system as the lateral loads will mostly cancel each other out. In which case I'd ensure your rafter system along the column line can handle the tension (which could be a challenge at the apex).

Also I'd also consider allowing for additional lateral load restraint which may be required due to uneven loading. (EG deeper snow on one side of the roof causing a resultant lateral load.)

 

[Ashke15]

Thank you for your feedback thus far everyone, it's very helpful.

SWComposites and human909, here is a drawing showing some more detail. I added in some fly braces similar to what both of you were talking about, at least I think it captures what you were talking about. I have them shown at every other rafter on my drawing.

I also drew a cross section of how the architect shows the rafters butting up against the trusses. I have no idea what his intention was here so I've got to come up with something that will work. One idea is to attach a ledger to the trusses and truss blocking. Another idea is to continue the rafters back into the structure and provide a short stud from the rafter down to the wall plates.

design heavier rafters about the columns if required

If I place the braces at every other rafter or at each rafter then it should disperse the horizontal load into all the rafters instead of just the ones at the ends, right?

 

[SWComposites]

The fly braces as shown are not going to be that effective for preventing rolling / lateral instability of the beams. I would have plates on the end of the braces that connect to the full height of the beam web.

And I would connect them to the rafters with more than a single bolt.

 

[human909]

SWComposites and human909, here is a drawing showing some more detail. I added in some fly braces similar to what both of you were talking about, at least I think it captures what you were talking about. I have them shown at every other rafter on my drawing.

That is what I would typically expect by fly braces. Whose typical role is to restrain the bottom flange of a rafter or similar member.

However in this case I was referring to fly braces SPECIFICALY at the locations of the columns sized to take the full column lateral load so no twist is induced in the beam itself.

If I place the braces at every other rafter or at each rafter then it should disperse the horizontal load into all the rafters instead of just the ones at the ends, right?

Possibly... But that wouldn't be my choice. That is a longer and complicated load path, it also relies on the minor axis of the beam to transfer lateral load. Keep it simple. Design it so there is a rafter above the column and design that rafter specifically to take and transfer ALL the lateral load from the column.

I'd only use braces at the rafter located at the columns unless they are required elsewhere for LTB. I'd clearly calculate the lateral load path all the way to the other column which presumably has an almost symmetric and opposite lateral load.

 

[Ashke15]

I'd clearly calculate the lateral load path all the way to the other column which presumably has an almost symmetric and opposite lateral load.

I'm not sure what you mean by this. The horizontal load at the top of the columns is due to the vertical load from the 25' long beam that bears on top of each column. So in my first drawing where I show a 2288 lb. horizontal load I believe that 2288 lb. will be present at each column. So I will need a rafter above each column that is capable of resisting 2288 lb. of force from the beam and column that is trying to pull the rafter away from the back wall of the house. I'm not understanding why the lateral load will be opposite. I probably described something incorrectly in one of my previous posts.

 

[BAretired]

I've got a large custom home in a high snow load area with a covered deck at the back. The deck is 25' wide and has an inclined steel column at each end. There will be a beam at the top to carry the roof and a beam at the deck level. I have resolved the loads into axial compression in the column, vertical reaction at the bottom, and horizontal reaction at top and bottom using the D + .75S + .75L load combination. My question concerns the horizontal reaction at the top of the column. What prevents the beam from rotating as it tries to resist the horizontal force from the column?

The beam of course welds to the top of the column. It appears to me that the column tries to rotate the beam away from the house. The welds hold the beam in place, but then the beam tries to pull the rafters away from the house. Is this the correct load path? [highlight #FCE94F]It is the only load path, but it should be a sloping beam, not rafters.[/highlight] I need to make sure that the rafters are secured appropriately to the beam and back at the house so that the beam doesn't pull away from the rafters, and so that the rafters don't pull away from the house. [highlight #F57900]You need to make sure that the house and the connection can withstand the horizontal force.[/highlight] That's how I see it, but wanted to see what you think. Do the deck joists see any of that force pulling them away from the house? Would appreciate any feedback. [highlight #8AE234]The deck joists will if they are tied in at both ends, but you should not rely on them.[/highlight] In summary:

1) The beam rotation is resisted by the welds from the beam to the column? I will also have web stiffeners.
Not good enough...continue column up and connect to sloping beam.
2) The beam imparts a horizontal force on the rafters pulling them away from the house? I would not rely on rafters. Connect the sloping beam for the full force.
3) Do the deck joists also pull away from the house? They could if they are securely fastened at each end. There is also wind pressure.
4) Recommendations from your experience? I have told you all I know already.
5) Is there a certain degree of incline where you need to design the column with a moment at the base instead of only vertical and horizontal reactions? Each case is different.
6) Where did I mess up? Who says you did?

 

[BAretired]

The deck is 25' wide and has an inclined steel column at each end.

Some of us were interpreting that statement incorrectly. Some thought you meant the structure was symmetrical from side to side rather than from end to end. If that were the case, the sloping columns would have horizontal reactions which tended to balance each other.

 

[Ashke15]

It is the only load path, but it should be a sloping beam, not rafters

Thank you for your response. I want to make sure I understand this part. Are you suggesting that instead of bearing the beam on top of the column that the column extends all the way to the bottom of the rafters and the beam hangs into the side of the column? And are you also suggesting that I rotate the beam so that the rafters have a flat surface to bear on? See below...

Who says you did?

No one, but when I come across something new I usually assume that I have no idea what I'm doing.

 

[human909]

I agree with BAretired. Your above picture seems a much more sensible approach to address the beam to column connection. Welding it to the web on site would likely be more expensive and unnecessary unless there are architectural requirements that make bolting undesirable.

I'm not sure what you mean by this.

I think I misunderstood your comment "it's symmetrical on the other side." Which BAretired pointed out.

So I will need a rafter above each column that is capable of resisting 2288 lb. of force from the beam and column that is trying to pull the rafter away from the back wall of the house.

YES. And I'd be designing the rafter to resit this load and providing a clear load path to take this load to your lateral restraint system. I wouldn't just blindly trust that 2288 lb load will just happily transfer through your roof diaphragm.

If it isn't clear a "clear load path" could be as simple as a few tensioned steel strap or rod braces from the rafter to a suitably stiff and laterally restrained point.

 

[BAretired]

Use a sloping beam connected to the house for the full horizontal load at each end. Use nailers for the rafters, and if rafters take part of the load, so be it, but don't count on it.

 

[Ashke15]

Awesome, thank you all for your input and drawings. This really helps a lot!

 

[LittleInch]

Is there a good reason why the column to rafter connection isn't at 90 degrees? This would surely eliminate the issue at source rather than trying to fix the rotation issue?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[human909]

Is there a good reason why the column to rafter connection isn't at 90 degrees? This would surely eliminate the issue at source rather than trying to fix the rotation issue?

I'm sure there is. I suppose there are less architects involved in your petroleum engineering world than in residential construction.

I'd also note that people aren't eager to have their home look like an oil refinery.

 

[Eng16080]

Perhaps it's been addressed in the comments above (I only skimmed), but also be sure the overall frame is stable. Looking at the frame model (by BAretired), I was initially thinking it's unstable if all the joints are pinned, but thinking about it more, I'm actually not sure. With the upper portion of the frame being a trapezoid, I think that may inherently add stability.

As far as the original question, I agree with BAretired's approach.

 

[BAretired]

I was initially thinking it's unstable if all the joints are pinned, but thinking about it more, I'm actually not sure.

Well Eng 16080, we cannot afford to guess. Member BE may be pinned to DF, but Joint E is not pinned in the column. Point C, attached to the house must be capable of resisting the horizontal and vertical load from member CD. If it is not, we are in serious trouble. If it is, CDF forms a three hinged arch which is inherently stable. Member BE adds a vertical load at each end, but does not participate in stability of the arch.