Rafter thrust and wall plate.

[amwc]

Hi
I am designing a raised tie truss/ collar truss by hand and I hope some one can help me.
Once I have resolved all the forces in the truss and know the horizontal and vertical forces at wall plate/ rafter feet level, how do I convert this into a deflection?
Is there an equation for getting this figure?
Thanks in advance
Andy

 

[BAretired]

Why are you getting a horizontal force at the wall plate under gravity load? The usual assumption is that the wall does not take any horizontal force because it can't. Under gravity load, the deflection at each wall is one half the strain in the bottom chord, unless there is some reason to believe one wall is stiffer than the other.

EDIT: The last comment is intended for trusses with a horizontal bottom chord. For scissor trusses, or other more complicated shapes, any deflection may be found by using more sophisticated methods which can be found in elementary structural text books.

 

[MIStructE_IRE]

Analyse as a pin and roller. The outward horizontal movement of your roller is the deflection of the wall (the amount the top of the wall is pushed out). This should generally be less than wall height / 300 to avoid cracking in masonry.

 

[phamENG]

It's a whole lot of work. The unit load method will give you what you're after. This website has a synopsis and brief example.

The example isn't the same as your situation, but it gives you enough to be able to adapt the method to other applications.

 

[oldestguy]

I'll throw in my usual comment. There is beam action in that plywood roof, in the plane of the roof. Takes most of the thrust.

 

[XR250]

I'll throw in my usual comment. There is beam action in that plywood roof, in the plane of the roof. Takes most of the thrust.

Where does it go?

 

[Harbringer]

Where does it go?

I think he means that it's taken through diaphragm action to the shear walls perpendicular to the truss span. Seems reasonable.

 

[oldestguy]

Liken this to a cardboard box of that shape. End walls may take it. So fold a piece of heavy paper with a peak, like a tent. Glue a triangular piece for each end. Put a load on the ridge. Now answer the question.

 

[phamENG]

I'm pretty sure the OP is not asking about how thrust is resisted, but rather how to calculate the deflection of a truss by hand, specifically the horizontal thrust at the heel.

 

[Harbringer]

I am designing a raised tie truss/ collar truss by hand

Ahh I missed the raised collar tie part.

To the OP, provide more information starting with a sketch. Are you designing a truss or rafters with collar ties? Are you trying to determine top of wall deflection OOP (important with collar ties) or just roof deflection?

 

[XR250]

Liken this to a cardboard box of that shape. End walls may take it. So fold a piece of heavy paper with a peak, like a tent. Glue a triangular piece for each end. Put a load on the ridge. Now answer the question.

I think he means that it's taken through diaphragm action to the shear walls perpendicular to the truss span. Seems reasonable.

Sounds good in theory and works fine on a tiny cardboard box model. Get down into the details of the analysis on a real-life, typically constructed roof and it does not (generally).
I have inspected over 6000 houses and can tell you from experience that "beam action" seems to work for smaller roofs but most of the time it does not. Long-term creep, nail slip and general overstress in the system due to the numbers not working out will cause the exterior walls to bow.

 

[XR250]

...actually 5000 houses, but who is counting

 

[RWW0002]

@XR250 - 5000 houses is alot of houses! Not to hijack the thread completely but I would be curious about your inspection experience in general. I live in an area with prescriptive-only residential design and, as many of us know, the adherence to prescriptive codes can be variable and highly dependent on local practice, local code officials & inspectors knowledge/whims, and the experience of contractors. I do not do much residential design, but when I do get pulled into a project I often find whole areas of the code that are overlooked by most or all partied involved, without the expected (lack of) performance. I sometimes wish I had a myriad of examples to discuss/show to keep from being seen as an alarmist all the time. When you have a minute I would love to see your "top 10" list of issues that you come across due to poorly designed/built houses.

Now back to the issue at hand - @ amwc One other approach I have used in the past in lieu of the pin-roller mentioned above is to envision the wall/roof framing as a simple (gravity only) frame. Include pinned-base columns corresponding to the wall height at each rafter support. This will allow each side of the truss to spread and should give you an idea of the magnitude of the wall deflection at the plate level. Ultimately this should be very similar to using 1/2 the value for the horizontal deflection of pin-roller truss, but I have found it useful for some slightly more complex situations.

 

[Harbringer]

...End walls may take it...

...general overstress in the system due to the numbers not working out...

Yes, inadequate or not even engineered roof framing may cause walls to bow. My guess is of the 6000 houses you've looked at a tiny fraction actually had an engineer involved (maybe aside from pre-engineered roof trusses) and most of the failures you've observed are due to contractor or DIY'er "designs" with the related poor detailing.

 

[RWW0002]

@Harbinger, Even properly designed trusses can cause wall bow, sometimes significant, if they are scissor trusses or some kind of raised collar vaulted truss as described above. I have looked at diaphragms carrying roof thrust before, and for a building of any size with scissor trusses or similar the load can get pretty large at the end shear walls. I have no doubt that the roof diaphragm helps reduce deflections in many cases, but it is not a magic bullet by any means.

 

[phamENG]

RWW0002 - quite right. I looked at one a few months ago - the scissor trusses were only present over the central portion of the house. So the "shear wall" on each end was just an interior partition with gyp, and that gyp was splitting apart right below the ridge line.