The IBC code is clear as long as there is a structural ridge beam with posts no thrust will occur at the exterior walls. Therefore no heel connection design is required. Is this because of plate action of the plywood. Any tips will be greatly appreciated. When I place the snow load perpendicular along the slope of the rafter and not along the horizontal thrust forces occur. This is easily verified with Ram Advanse.
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
Structural Ridge Rafter Thrust 8
- Thread starter cap4000
- Start date
woodman88
Structural
- Oct 23, 2009
- 1,020
- 0
- 0
BAretired - I know this and agree with it. The point I am trying to make is that the "horizontal projection method" does not calculate this within the method.
It is there and you can calculate it as you have done. But because it is not in the method the method is technically incorrect in that it does not do all the calculations required.
So I must agree with Breyers book Design of Wood Structures statement and the fact that the method is "....typically used but technically incorrect."
Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.
It is there and you can calculate it as you have done. But because it is not in the method the method is technically incorrect in that it does not do all the calculations required.
So I must agree with Breyers book Design of Wood Structures statement and the fact that the method is "....typically used but technically incorrect."
Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.
woodman88,
I wasn't even aware that it was called "the horizontal projection method". It is just plain statics and it can be done by either method.
I disagree that either method is technically incorrect simply because it does not perform all of the calculations. You might want to say it is incomplete, but why would you say in is incorrect?
BA
I wasn't even aware that it was called "the horizontal projection method". It is just plain statics and it can be done by either method.
I disagree that either method is technically incorrect simply because it does not perform all of the calculations. You might want to say it is incomplete, but why would you say in is incorrect?
BA
This thread is very fun. However I think the OP needs to supply some FBD w/loading of the different analyzed scenarios in order to adequate discuss this. Pictures would be very nice too. It seems to me that everyone here are very experienced and know exactly what to do but are not connecting at the same points.
Also from earlier post on the pending investigation, I personally have lots of pictures of partially withdrawn nailed members that have NOT failed to collapse. So, not saying that this isn't the cause but it certainly doesn't necessarily follow that that it was the cause of the collapse based on the information given... Just saying.
Further, in a mono slope situation a lot of lateral/horizontal deflection could occur at the ridge which typically would be balanced and developed to the eave walls in a typical gable.
In short, my $.02 on this: Upper support on the rafter in the vertical direction should contain the rafter "thrust" to negligible as long as:
1) the ideal rafter does not deflect too much
2) the ridge support does not deflect too much vertically
3) lateral/horizontal forces do not enter into the model(snow is a vertical load globally) and no horizontal deflections of the supporting frame were taking place.
4) the connection at the heel can be idealized as a horizontal roller.
5) the connection at the ridge can be idealized as a pin.
Since we don't have any specifics on the situation It is safe to say: "there might be thrust and there might not".
______________
MAP
Also from earlier post on the pending investigation, I personally have lots of pictures of partially withdrawn nailed members that have NOT failed to collapse. So, not saying that this isn't the cause but it certainly doesn't necessarily follow that that it was the cause of the collapse based on the information given... Just saying.
Further, in a mono slope situation a lot of lateral/horizontal deflection could occur at the ridge which typically would be balanced and developed to the eave walls in a typical gable.
In short, my $.02 on this: Upper support on the rafter in the vertical direction should contain the rafter "thrust" to negligible as long as:
1) the ideal rafter does not deflect too much
2) the ridge support does not deflect too much vertically
3) lateral/horizontal forces do not enter into the model(snow is a vertical load globally) and no horizontal deflections of the supporting frame were taking place.
4) the connection at the heel can be idealized as a horizontal roller.
5) the connection at the ridge can be idealized as a pin.
Since we don't have any specifics on the situation It is safe to say: "there might be thrust and there might not".
______________
MAP
woodman88
Structural
- Oct 23, 2009
- 1,020
- 0
- 0
BAretired - Because you should say "something" to made people aware that there may be potential issues with assuming it is completely correct?
Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.
Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.
It is completely correct for the assumptions made, i.e. hinge at ridge and horizontal roller at wall. The load and reactions are vertical and may be resolved into components parallel and perpendicular to the rafter or, for that matter, any other directions desired.woodman88 said:
BA
-
1
- #46
jimodore
Structural
- Jul 26, 2012
- 6
- 0
- 0
Hello to all from the UK where the more frequent variation of this problem is the horizontal spreading of the top of masonry walls supporting sloping rafters without horizontal ceiling joists. The movement results in cracks in the horizontal mortar joints at the top of the walls.
So, here is my two pence woth:
If the support from a wall or posts to the bottom of the rafters effectively allows free horizontal movement, then the reactions there will be vertical, whilst at the ridge the reactions will have vertical and horizontal components. An effective diaphragm in the plane of the roof would transmit the horizontal load component to the gables, reducing or eliminating the horizontal component at the ridge.
However, this (pin and roller supports) model is unlikely to be the realistic in practice if the wall or posts resist horizontal movement to any degree. Timber posts may readily accommodate the movement by flexing whereas masonry walls tend to suffer horizontal cracks in the upper mortar bed joints - but there will be some restraint foces in both cases. This does not detract from using this modelling for the sizing / design of the timber sections.
A more useful model to illustrate the potential for horizontal restraint forces at the wall / post top is to consider the rafter as pinned at both ends, with the posts / wall either pinned or fixed at their feet. Such modelling gives an upperbound limit on what forces might be generated if the joint between the rafters and the wall/posts resists movement to some degree. As noted above, an effective roof diaphragm will reduce or eliminate such hoizontal reaction forces. This modelling approach is not advocated for the rafter sizing because of the extra complexity in the analysis and the uncertainty around the degree of restraint at the top of the posts / wall.
Footnotes:
1. Other mechanisms for generating horizontal forces at the post/wall to rafter connection include:
a) pressure / suction on the elevation (depending on the solidity)
b) horizontal components of wind loads on the roof
c) the spreading of the soffit of the rafter under vertical load due to tensile strains.
2. Memo to self: I must keep things short and sweet......
So, here is my two pence woth:
If the support from a wall or posts to the bottom of the rafters effectively allows free horizontal movement, then the reactions there will be vertical, whilst at the ridge the reactions will have vertical and horizontal components. An effective diaphragm in the plane of the roof would transmit the horizontal load component to the gables, reducing or eliminating the horizontal component at the ridge.
However, this (pin and roller supports) model is unlikely to be the realistic in practice if the wall or posts resist horizontal movement to any degree. Timber posts may readily accommodate the movement by flexing whereas masonry walls tend to suffer horizontal cracks in the upper mortar bed joints - but there will be some restraint foces in both cases. This does not detract from using this modelling for the sizing / design of the timber sections.
A more useful model to illustrate the potential for horizontal restraint forces at the wall / post top is to consider the rafter as pinned at both ends, with the posts / wall either pinned or fixed at their feet. Such modelling gives an upperbound limit on what forces might be generated if the joint between the rafters and the wall/posts resists movement to some degree. As noted above, an effective roof diaphragm will reduce or eliminate such hoizontal reaction forces. This modelling approach is not advocated for the rafter sizing because of the extra complexity in the analysis and the uncertainty around the degree of restraint at the top of the posts / wall.
Footnotes:
1. Other mechanisms for generating horizontal forces at the post/wall to rafter connection include:
a) pressure / suction on the elevation (depending on the solidity)
b) horizontal components of wind loads on the roof
c) the spreading of the soffit of the rafter under vertical load due to tensile strains.
2. Memo to self: I must keep things short and sweet......
msquared48
Structural
BA:
I agree with you on your last post, however, I do not feel that the model stated is "absolutely" correct here, as the "roller" is not frictionless in reality at the wall.
There is a connection that is required by code at the ends of the rafters, and this does allow a lateral force, not only to be resisted (wind), but also to be generated, as in the small amount of thrust generated in the deflections of the components.
Mike McCann
MMC Engineering
I agree with you on your last post, however, I do not feel that the model stated is "absolutely" correct here, as the "roller" is not frictionless in reality at the wall.
There is a connection that is required by code at the ends of the rafters, and this does allow a lateral force, not only to be resisted (wind), but also to be generated, as in the small amount of thrust generated in the deflections of the components.
Mike McCann
MMC Engineering
jimodore said:
Elementary statics prevents having gravity load, a vertical reaction at the wall and horizontal components at the ridge. That is just plain wrong. The ridge reaction must also be vertical.
If the walls resist lateral movement, then I agree that if the ridge beam deflects even slightly, there will be horizontal forces. But the horizontal force is the same at the ridge as it is at the wall.
BA
New Postmsquared48 (Structural)
30 Jul 12 16:07
BA:
I am not suggesting the model is absolutely correct, Mike. But the first post made it clear that we are considering no lateral force on the wall. All of my comments have been based on that assumption.
For a wood stud wall, I would assume that is substantially correct. For a masonry wall, probably not.
If there is a ridge board but no ridge beam, then as you mentioned earlier, collar ties would be needed.
If there is no ridge beam and no collar ties, the roof can still carry snow load as a folded plate but that entails careful detailing of the roof diaphragm.
BA
30 Jul 12 16:07
BA:
msquared said:
I am not suggesting the model is absolutely correct, Mike. But the first post made it clear that we are considering no lateral force on the wall. All of my comments have been based on that assumption.
For a wood stud wall, I would assume that is substantially correct. For a masonry wall, probably not.
If there is a ridge board but no ridge beam, then as you mentioned earlier, collar ties would be needed.
If there is no ridge beam and no collar ties, the roof can still carry snow load as a folded plate but that entails careful detailing of the roof diaphragm.
BA
woodman88
Structural
- Oct 23, 2009
- 1,020
- 0
- 0
jimodore - You did not mention deflection as a cause for horizontal reactions in a pinned pinned rafter/beam. You know, the shortest distance between two points is a straight line. Force the line to curve and the distance between the two points must get smaller or the line longer. So to keep the two points in the same location and not allow the line to get longer the line must be pulled back to a straight position. Creating a horizontal reaction at the pinned supports.
Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.
Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.
Why would that be, jimodore?jimodore said:
Would it not be more correct to say that the cracks occurred at the bottom of the walls where the moment is maximum?
jimodore said:
A much more useful model is to assume that the wall has no resistance to horizontal force and that the rafter is required to take the full moment without benefit of horizontal resistance from the wall.
woodman88 said:
Why would he mention that, woodman88? A pinned/pinned situation is not representative of a real structure.
Perhaps the best thing that could happen to this thread is that it be removed from Eng-Tips because it demonstrates that practicing engineers do not understand the true nature of the structures they are designing.
Alternatively, perhaps it should remain in Eng-Tips so that young engineers could study the problem at length and determine whose brand of baloney they are prepared to buy.
BA
woodman88
Structural
- Oct 23, 2009
- 1,020
- 0
- 0
BAretired - Actually the pinned pinned bearing condition is one of my pet peeves in wood structures. It can occur in two basic conditions. The first is at scissor trusses at the gable end of a building. The second is with scissor trusses (or roof/ceiling rafters to a ridge beam) at the midspan of a exterior wall with perpendicular interior non-bearing walls connecting to them.
In both conditions the scissor trusses want to push the exterior wall away from the perpendicular walls (either the gable end or the interior non-bearing walls) or have their bearing slide out over the exterior wall. Usually in the most extreme cases this is only 1/4” to 1/2” at each end. This rarely occurs at the gable ends as the roof diaphragm usually will resist this horizontal deflection, but I doubt if it is being calculated for this. But I have heard (no doubt only because of my 15 years in the wood truss industry) complaints about this and the answer is that it is not going to move much now so patch the wall.
It is a rare situation that only occurs with a sloped ceiling mainly in custom homes and unless you had had the problem you probably do not care. But I wished that more engineers or building designers doing these type of structures would as least take a closer look at it.
As for you comment "Perhaps the best thing that could happen to this thread is that it be removed from Eng-Tips because it demonstrates that practicing engineers do not understand the true nature of the structures they are designing." I worry more about the structures retired engineers have done.
Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.
In both conditions the scissor trusses want to push the exterior wall away from the perpendicular walls (either the gable end or the interior non-bearing walls) or have their bearing slide out over the exterior wall. Usually in the most extreme cases this is only 1/4” to 1/2” at each end. This rarely occurs at the gable ends as the roof diaphragm usually will resist this horizontal deflection, but I doubt if it is being calculated for this. But I have heard (no doubt only because of my 15 years in the wood truss industry) complaints about this and the answer is that it is not going to move much now so patch the wall.
It is a rare situation that only occurs with a sloped ceiling mainly in custom homes and unless you had had the problem you probably do not care. But I wished that more engineers or building designers doing these type of structures would as least take a closer look at it.
As for you comment "Perhaps the best thing that could happen to this thread is that it be removed from Eng-Tips because it demonstrates that practicing engineers do not understand the true nature of the structures they are designing." I worry more about the structures retired engineers have done.
Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.
Woodman88,
Your concern about structures designed by retired engineers may be valid. I have seen a lot of bad design over the years. But this thread is a clear indication to me that some of the current engineers do not understand simple statics. To me, that is a huge concern.
BA
Your concern about structures designed by retired engineers may be valid. I have seen a lot of bad design over the years. But this thread is a clear indication to me that some of the current engineers do not understand simple statics. To me, that is a huge concern.
BA
- Thread starter
- #54
BA
Take a look at the link below when you catch your breath. This thread is about a snow load perpendicular to the beam with a Fx reaction.
Take a look at the link below when you catch your breath. This thread is about a snow load perpendicular to the beam with a Fx reaction.
Cap - That link proves BA's point there is NO Fx reaction (Fx=Ha=0lbs) for Gravity loads (dead and snow) only for wind load is there a Fx reaction that is not equal to 0.
EIT
EIT
-
1
- #56
woodman88
Structural
- Oct 23, 2009
- 1,020
- 0
- 0
I think the main point that is being missed by most of the comments here is that the reason the sloping beam/rafter does not move is because we are assuming the bearings are/will be horizontal. This assumptionn of horizontal bearing may be valid in 99% or more of the situations and when one specifies a pre-made sloping hanger, a Simpson hanger for instance, the axial force parallel to the rafter/beam has been considered for us. Whether by the hanger designer or the testing.
The huge concern to me, is that with so many of the comments here are assuming a horizontal bearing. But continuing to merely to restate "no horizontal reaction" and not adding that this is only because of the assumed horizontal bearing.
Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.
The huge concern to me, is that with so many of the comments here are assuming a horizontal bearing. But continuing to merely to restate "no horizontal reaction" and not adding that this is only because of the assumed horizontal bearing.
Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.
So for the sake of the argument:
What is the horizontal "thrust" of a rafter at a 45deg angle spanning 10 horizontal with a 20plf snow load (projected horizontally) and supported at the top by a pinned connection and a semi-pinned OR roller in the horizontal only? Assume no wind or seismic loading and dead load is non existant. Oh yeah, it should be a 2x12 rafter #2 DFL.
Values only! Do your work using statics only. Fancy schmancy software packaged answers not accepted and will result in an "F". We have enough commments already. Put your money where your mouth is. FEA matrix analysis will earn you extra credit plus you would have to do this if you assume a pinned connection at the bottom.
______________
MAP
What is the horizontal "thrust" of a rafter at a 45deg angle spanning 10 horizontal with a 20plf snow load (projected horizontally) and supported at the top by a pinned connection and a semi-pinned OR roller in the horizontal only? Assume no wind or seismic loading and dead load is non existant. Oh yeah, it should be a 2x12 rafter #2 DFL.
Values only! Do your work using statics only. Fancy schmancy software packaged answers not accepted and will result in an "F". We have enough commments already. Put your money where your mouth is. FEA matrix analysis will earn you extra credit plus you would have to do this if you assume a pinned connection at the bottom.
______________
MAP
I do not believe we are assuming the bearings are horizontal. They might be, but do not have to be.
I am assuming the wall is incapable of resisting a horizontal force, i.e. hinged top and bottom. For a wood stud wall that is not far from the truth. For a concrete or masonry wall, perhaps not. If the wall is capable of resisting only a vertical force, there can be no horizontal reaction at the low end of the rafter. The connection between rafter and wall must be designed to accommodate the reaction.
The connection at the ridge beam may be a joist hanger with sloping seat, but if so, the nails must be sufficient to safely sustain the axial tension at that end.
BA
I am assuming the wall is incapable of resisting a horizontal force, i.e. hinged top and bottom. For a wood stud wall that is not far from the truth. For a concrete or masonry wall, perhaps not. If the wall is capable of resisting only a vertical force, there can be no horizontal reaction at the low end of the rafter. The connection between rafter and wall must be designed to accommodate the reaction.
The connection at the ridge beam may be a joist hanger with sloping seat, but if so, the nails must be sufficient to safely sustain the axial tension at that end.
BA
- Status
Similar threads
- Question
- Question
