×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

Roof Live Load

Roof Live Load

Roof Live Load

(OP)
I don't work in this stuff everyday, so please forgive me for the elementary nature of this question.  ARE ROOF LIVE LOADS APPLIED VERTICALLY (GLOBAL -Y) TO SLOPED ROOF MEMBERS?  (i.e. not perpendicular to sloped member?)  Seems like they should bc big part of roof live is weight (i.e. vertical load) of various construction incidentals.  

However, I get confused bc codes/stds/packages make point of distinguishing horizontal projection vs real length.  If it's vertically applied, isn't it always the horz projection?  Thanks.

RE: Roof Live Load

Might not be a single answer to your question.
If you look at it simply as the direction that loads are acting, then gravity loads are globally vertical and downward.
Pressures act normal to the surface.
so, say for a purlin design on a sloped roof, gravity loads would cause bi-axial bending in the purlin, but pressure loads would cause only uni-axial bending.

agreed?

RE: Roof Live Load

(OP)
Agreed.  This case is a truss w/ 20psf specified LL on top chord.  Only looking at truss, no purlin w/ biaxial.  Just trying to tune a RAM Elements model to match reactions, etc provided on output from a MiTek model.  Thanks for chiming in.

RE: Roof Live Load

If someone chimes in about shear centers and purlins that are channel shapes, I will go home immediately and start the Holiday weekend early!!!!

RE: Roof Live Load

Dont forget the torsion induced by...darn ToadJones has beeten me to it :)

RE: Roof Live Load

csd,
don't do it man, I'm serious!
If BARetired starts making diagrams about torsion, I'm outta here!!!!

RE: Roof Live Load

The answer to the actual question is per IBC 1607.11 Live and Snow loads both act vertically on the horizontal projection of the beam.  

You would multiply the live load by cosine of the rise to run angle to get the vertical load acting along the slope.

You multiply that value by cosine of the rise to run angle to get the component of load acting perpendicular to the member, or by sine of the angle to get the component acting parallel (axial force) to the member.

So...in summary:

Component perpendicular to member:  LL(cos theta)^2
Component parallel to member:  LL(cos theta)(sine theta).

where theta = rise to run angle in degrees.   

RE: Roof Live Load

So....
Say you have:

A truss with a 4:12 pitch, a 10 ft tributary width, and live load of 20 psf.

Say you want to model this truss with a uniform distributed load acting in the Global Y (down) direction, then you are saying that load would be:

rise = 4'
run = 12'
hypot.= 12.65'

vertical load (12/12.65)* 20psf * 10' = 189.7 plf?

And, from here you'd break that into forces perpendicular and parallel to the top chord?  

RE: Roof Live Load

ToadJones - correct - you would take the 189.7 and multiply it by cos or sin for perpendicular and parallel components.

RE: Roof Live Load

hmmmm...
I honestly never interpreted in that way.
I guess it is a good thing most of the roofs I have done are 4:12 and under.
Even at 4:12 it seems to only represent about a 6% change.  

RE: Roof Live Load

hubba,
For a detailed examples of the Sloping Beam and Horizontal Plane Methods of analysis see pages 2.19 to 2.21 of the 6th edition of Design of Wood Structures (ASD/LFRD) by Donald E. Breyer, et al.
Basically the two approaches yield the same results (monents & shear).

RE: Roof Live Load

This is an east mistake to make in specifying the load in an analysis program. You need to pay attention to the defaults and make sure you apply it locally or globally on purpose. RISA3D for instance can do it either way, you just need to specify it properly.
 

RE: Roof Live Load

In higher pitched roofs, it does make a difference in the rafter size.  

It is particularly applicable in Hip and Valley member design.

Mike McCann
MMC Engineering
Motto:  KISS
Motivation:  Don't ask

RE: Roof Live Load

If you have a truss
with a 4:12 pitch and 10' trib width with 20 psf LL why wouldn't the line load simply be 20 psf * 10' = 200plf ; then use that to determine the load acting perpendicular to the beam (causing bending) and in line with the beam (causing axial load)?  The 10' trib isn't sloped so the horizontal projection is 10'.

RE: Roof Live Load

SEIT - LL is specifically given by code to act on the horizontal projection of the beam, so the 20psf has to be placed along the slope of the beam before you can calculate the parallel and perpendicular components.  What you stated is correct for dead load etc. (loads that are already along the length of the beam don't need that initial multiplying by cosine step from my above post).   

RE: Roof Live Load

What am I missing? Isn't the LL already acting vertically? Why do you need to do anything to it to get it to act vertically?  I see the part about taking the components perpendicular to and parallel to the beam, but I'm missing something when it comes to the acting vertical part.

If you have a beam that spans horizontally 10' with a trib width of 10' and a LL of 20psf and is flat, the vertical reaction at each end is 1000#.  Of you keep the horizontal span of 10', but offset the ends vertically by 3', the  vertical reaction at each end is still 1000#, no? In both cases the horizontal projection is 10'.

RE: Roof Live Load

The internal forces will be different for each, because they have different lengths and different components acting perpendicular to and parallel to the axis of the beam, but in both cases the vertical reaction is the same.

RE: Roof Live Load

SEIT yes, you are missing something - it's not the vertical term that is key, it's that it needs to be vertical along the length of the beam because the code gives you the live load values in terms of horizontal projections - I'll post a diagram next week when I can - basically for a 45deg roof the code roof horizontal projection LL value of 20 psf results in an actual load of 14.14plf acting vertically on a slope along the beam.  
You can get correct moments without dealing with this, but this is the only way to get correct axial forces.    

RE: Roof Live Load

I'm with StructuralEIT.  Snow is a gravity load and is specified by code on a horizontally projected area.  If it is 20 psf, then the total snow load acting on plan area A is 20*A, irrespective of roof slope (unless it is steep enough for the snow to slide off).  

Dead load is also a gravity load, but if the dead weight of deck plus roofing materials weighs 'w' psf and the slope is θ then the dead weight on plan area A is w*A/cosθ.

BA

RE: Roof Live Load

Toad,

Sorry to disappoint you.  No diagrams about torsion.

BA

RE: Roof Live Load

Careful BA.  He gets torqued easily.

Mike McCann
MMC Engineering
Motto:  KISS
Motivation:  Don't ask

RE: Roof Live Load

Just a side question, but under most circumstances doesn't a sloped roof member need to be sheathed in which case the longitudinal force goes into the diaphragm, not as an axial force in the member,  right?  

RE: Roof Live Load

It can, yes.  

But if there are rafters each side of a ridge beam, and a the rafters are tied to the ridge in tension, the longitudinal force is never seen, either to the purlin or the diaphragm.

Mike McCann
MMC Engineering
Motto:  KISS
Motivation:  Don't ask

RE: Roof Live Load

A rafter with a birdsmouth joint at the wall plate and a joist hanger at the ridge beam has a vertical reaction at each end.  There are no horizontal forces involved, so the bending moment is the same as a horizontal member of the same span.  Reactions can be resolved into components parallel and perpendicular to the rafters to find shear and axial loads.  I wouldn't say that the axial load goes into the sheathing, although some might.

Tied rafters do not require a ridge beam.  Often, a ridge plate is placed between opposing rafters.  Rafters can have a birdsmouth joint at the walls.  Ceiling joists must be continuous across the entire span and must be spliced to carry tension from wall to wall.  The rafters feel more compression in this case and some of that compression may be transferred to the sheathing, but it may not be prudent to rely on that.

Sometimes, rafters are tied half way up the rafter.  This changes the statics considerably.   

BA

RE: Roof Live Load

I think the confusion here is mostly just semantics.
The code simply states that the live load shall be taken as the horizontal projection.
After some thought, it makes sense. Dead load is there, can't do anything about that. But in the case of a steep roof (say 12:12) it would certainly be more difficult to "load the roof" with live loads especially during construction.
Thoughts?

RE: Roof Live Load

Don't overthink it.  Gravity loads always act down (-Y).  They have to, the earth is pulling them that way.

Roof live loads and snow act downward on the projected area.  Dead loads also act downward, but you need to apply the whole sloped area because that is all actual material with weight.  If you've got a 45-degree roof with 10' of rise and 10' of run, on the rafters you would apply [plf weight]*[2/sqrt(2)] on the horizontal projected area.

RE: Roof Live Load

Hubba:
MiTek and others hold their cards pretty close to their vest when it comes to their design and analysis programs, and it is becoming more and more difficult to do some simple calcs. to verify their controlling reactions and member stresses and sizing, etc.  What with all the load combinations, variable wind loadings and nonuniform snow and drift loading, and then add earthquake loadings, and who knows but that a different load combination doesn't control every single member in the truss.  I expect that, in another generation or so of code changes a simple building will be impossible to design and analyze by what we used to call rational engineering methods.  You won't ever get past trying to summerize the potential different load combinations, let alone know how to factor them up or down, and in wood, don't forget all those adjustment factors also.  And, then if you use different programs, you most likely will get different answers, because the various programmers interpreted an un-interpretable bunch of probabilistic babel differently, and none of them is really wrong; but their solutions are maybe more exacter (?) than  slide rule solutions were 50 years ago, they just need sixty more pages of printout to get there.  And, you needed more time to compile the load combinations now than I did to design the whole roof system.  They have ten decimal place answers, that must be gooder, but my building is still standing after 50 years, so who is righter?  If that roof structure knew what contortions you and they went through to prove it might stand up under load, it would fall down from sheer exhaustion or sheer exasperation.  That's different than horiz. shear stress, but is going to be included in the next version of the NDS.  And, LEED says it's greener too, when you use more trees to produce the paper for the paperwork and computer printouts than you do to produce the actual roof trusses.

RE: Roof Live Load

Finally, I think I see what WillisV is saying.  If the snow load is 20 psf, the angle is θ and the spacing is b, then the load is 20*b per foot of horizontal span, but it is only 20*b*cosθ per foot of slope length.

If the dead load of the deck and roofing is 10 psf, then the load is 10*b per foot of slope length and 10*b/cosθ per foot of span.

BA

RE: Roof Live Load

TJ:
Well.....  the confusion is actually a good deal more than just semantics.  The designer/engineer has to be smart enough to know when he/she needs to finesse the problem or when they can use the simplification implied by the code and our common practice for normal slopes.  The confusion has more to do with the proper interpretation of vertical loads ON a horiz. projection (not, "taken as the horiz. projection"), vs. load components perpendicular (bending) and/or parallel (axial) to the axis of the rafter; and the rafter's horiz. projected length and/or its actual sloped length.  You correctly saw that for normal pitch roofs the correction (or error) isn't too great, only about 6% for a 4:12 pitch.  We understand how to deal with floor joists where the DL and LL are both vertical loads (globally) which also happen to be perpendicular to the axis of the beam and the axial loading is zero.  That's right out of our Strength of Materials understanding of beams.  I believe that for normal pitches if you use all loads ON a horiz. projection and the horiz. projected length of the beam to calc. moments and shears you get the same results as if using the sloped length of the beam and the loads converted to components perpendicular to and parallel to the sloped axis of the beam, they've both been converted appropriately by the cos  of the slope angle.  Except the math is tedious so we usually avoid it, but it does work out to give the correct reactions, etc., as long as you don't screw up the trig. conversions, but then we never know what to do with that axial loading since it's only 6%.

But, what happens when you have a 12:12 pitch (45° slope) or 20.78:12 pitch (60° slope) roof.  The snow load which is based on local ground snow loading, on a horiz. projection, slides off the roof; but the DL hangs on the rafter through its connection to the rafter, and the roof diaphragm, and the rafter starts acting as a column member with a fairly small bending component from these loads.  Now you are designing a completely different member, not a beam, but a beam/column which might buckle and isn't even braced along the tension flange.  The point is that at some point we are designing a column type member rather than predominantly a beam type member.  On heavy enough DL's you would want to pay attention to this cos  conversion, it's important, as WillisV and BA suggest.  Fact is, we add up the DL's wrt the plane of the roof and then generally apply them directly to the horiz. projection, but I'm not sure I know my loads within +/- 6%.

Long experienced framers seem to know what to do with hips and valley, through years of trial and error (failures).  Engineers scratch their heads for days trying to do an exact analysis and finally say close enough, go ahead and nail er down.
 

RE: Roof Live Load

BA-

I think I see the point, too.

RE: Roof Live Load

BAretired - bingo - glad you and SEIT have seen the light =)

RE: Roof Live Load

WillisV,

Sometimes my brain takes a few days to kick in.

BA

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources