rafter to ceiling joist connection

huangyhg

rafter to ceiling joist connection
i recently realized that i may not understand rafter to ceiling joist connections as well as i thought.  i always thought that the rafters sat on the top plate of the wall either with a bird mouth cut or with special brackets and that the ceiling joists were nailed to the rafters (with no rim joists running down the ends of the ceiling joists.
  please see the attached pdf.  it has blown my mind... i didn't know that this was legal.  how can this knee wall resist the rafter thrust?  when trying to find this answer i only came across more examples of people doing things similar to this.  
  i appreciate your help.
steven wheeler
a very strange detail indeed.  probably drawn by an architect or draftsman with very little understanding of how structures work.
i don't think you're far off... i've seen this type of detail used, correctly in my opinion, where the rafter sits on a ridge beam at the apex. the trouble comes when you have a ridge board (thus the apex deflects and gives rise to a horizontal thrust at the knee).
in no event do i think this detail is at all good, i'm just saying it's not always a red flag.  probably best not to use it just to minimize the chances of it being misused.
oh, and i have been asked what's wrong with using this with ridge board situations since it "works anyways"... my reply was that internal return walls, and even perhaps some stressed-skin action in the wall covering might make this okay to just build and admire on a nice day, but it will not take design loading.
cheers,
ys
b.eng (carleton)
working in new zealand, thinking of my snow covered home...
irc 2000 specifies that ceiling joists and rafters shall be nailed to each other and then references a table which gives the specifics for the connection.  recently had a project where the ceiling joists were running perpendicular and had to add rafter ties at 4'-0" oc.
although, i would bet the shown detail works, i'd prefer to keep the load path simple for residential construction.
hokie66,
this is the type of reason why i dont think architects should be allowed to design the structures for houses.
the enginnering for houses as actually often more complicated than many other types of buildings.
two things here:
1.  if there is a ridge beam, then either there is no lateral kick, or the kick can be taken out at the ridge beam connection with a crossover strap.
2.  if there is not a ridge beam, then the floor joists should be extended, considering the soffet detail so they can be nailed off to the rafters, thus serving as collar ties.
mike mccann
mmc engineering
the keyway drawn at the junction of footing/wall is of no structural use and adds to the cost.
the small pony wall at the top of the concrete wall supporting the floor joists should be concrete.  as it is detailed now, it's a great location for rot to start.
also, concerning the roof detail, the lateral spreading of the rafters can be resisted by using collar ties, and varying the location of the ties and the size and spacing of the rafters to control the lateral spreading to an acceptable level.
mike mccann
mmc engineering
the rafter detail problem is minor compared to the basement wall which is not braced by the floor and acting as a canatilevered retaining wall.  only problem is that the wall is not designed to be a retaining wall.
as far as the rafter detail, it can be easly fixed by adding a 2x4 kicker lapped to the base of the rafter and the attic joist or by baloon framing the wall.  a collar tie that is located 8 feet above the attic floor would not be an appropriate solution as the tied rafter bending stress and deflection would be to high.
i want to see a framer build the 5 1/4" knee wall (as drawn).
this section is poorly detailed all around. also note "dimensions" is misspelled and i've never seen a roof slope called out as 12:5.

did this come from one of those "find the errors" quizzes?