Collar Tie Design 2

[easleyml]

Can anyone provide a design resource for collar tie design for rafters. I have a contractor that has deviated from the requirements of the building code. Thanks

 

[rgerk]

Collar ties should just be loaded axially. Determine the axial load and design as a column, pinned at each end.

 

[Berniedog]

Do not know of any resources for design. Rule of thumb, no more than 1/3 up from rafter bearing at wall. We design simply by the following:
1. Solve for reaction of the rafter at the wall.
2. Solve for horizontal force in collar tie by method of sections. Create free body diagragm, of half the roof and solve about the ridge. Known values are the roof load and the reaction. Solve for horizontal force in collar tie.

Design connection to resist horizontal force.

 

[SlideRuleEra]

regerk and BrianRost have told you what you need to know, but if you require a short statement from a printed document, take a look at Paragraph 307.1.4 in Chapter 3 of "1999 SBCCI Standard for Hurricane Resistant Residential Construction" at this link (free .pdf download)

http://www.iccsafe.org/cs/standards/is-hrc/SSTD10/index.html

307.1.4 Collar Beams: A 1x6 collar beam shall be nailed in the upper third of the roof to every third pair of rafters.

http://www.SlideRuleEra.net

 

[HouseBoy]

SlideRule Era:
What do you suppose a collar beam of the sort described in the note, actually does?

Surely it has extremely limited structural function.

Your thoughts?

 

[SlideRuleEra]

houseguy - Good question. ASCE has published an paper, "Guidelines for Use of Collar Beams". I have not read it, but the abstract does answer that question

...in regions of moderate or heavy snow, collars can effectively prevent rafter separation at the ridge; in regions of high wind and no snow, either metal ridge straps or collars placed directly at the ridge should be used for most roof configurations. When used as rafter braces to permit larger spans, collar beams can have a modest impact, but only within a limited range of roof angles.

Here is the link to the paper

http://www.pubs.asce.org/

http://WWWdisplay.cgi?9700358

I assume that the Forest Products Lab determined recommended size and placement years ago since a summary of the use of collar beams in in their classic book "Wood Frame House Construction, Handbook No. 73". You can get a .pdf copy of this document from my website, see pages 78, 89, 223, 232, 242 & 248.

http://www.SlideRuleEra.net

 

[ctcray]

Is the OP asking about "collar ties" or "rafter ties"? Collar ties being in the upper third of the rafter to resist rafter separation at the ridge and unbalanced loading and Rafter ties being in the lower third resisting rafter/wall spread.

They work the same way, but serve different purposes. Collar ties aren't going to prenvent much rafter/wall spread.

 

[HouseBoy]

easleyme;
Often collar beam/ties actually perform very little structural function or at least there are multiple ways to accomplish the same thing.
How has the contractor deviated and what specific requirements?
It has been my experience that collar beams and ties are often misunderstood or their function is misstaded or stated in language that is easily misunderstood. If you could describe the conditions (or provide a sketch or link to illustrate things) perhaps that would be helpful.

SlideRuleEra;
Great post. Thanks

 

[rgerk]

I agree that collar ties serve a limited function. Unfortunately, the building community does not understand them and the rising complexity of the residential market is causing some builders, engineers, and architects to rely on collar ties to prevent the roof rafters from spreading at the wall plate.

I just had a conversation with a client for this very reason. They are trying to elevate the ceiling by using an LVL to span between rafters. The roof rafters are in place with a 2x10 (non-structural) ridge beam and there exists no tie between the top plates of the wall. I explained that the top of the wall has to be braced or the wall must be able to withstand the thrust force from the rafters. The other option would be to install a structural ridge beam.

Unfortunately, this isn't the first time this discussion has occurred with this client, but I believe he is finally starting to understand.

As BrainRost, stated, "no more than 1/3 up from rafter bearing at wall". Although, you should still consider deflection. Even if the rafter tie reduces the thrust it can still produce undesirable wall deflection. Thus leading to calls from the owner about drywall separation at the ceiling/wall corner.

Just food for thought.

 

[HouseBoy]

My experience regarding the one third up the rafter height "rule of thumb" is that quite often the bending in the rafter is too great (in addition to the "spread" you refer to).
The "rule of thumb" may function as a starting point (to begin to investigate the viability) but the tension connection in the tie, bending in the rafter member and deflection (actually both vertical deflection and horizontal spread) are often critical in such cases especially as the tie location gets closser to the 1/3 point.
It's not unusual for the rule of thumb to not "work".
Sometimes I can get adjacent floors to resist the thrust. Thsi may work unless the room projects out from the main body of the house. Also, if one side is not braced by a floor, all of the thrust will go out to the "open" side wall.

 

[dik]

Check the link

http://www.framingdesign.com/structural_engineering_framing_rafter.html

Dik

 

[HauerEng]

My suggestion is to determine the forces in the members using a simple finite analysis computer program. I think you will find, (particularly the higher to the ridge board the collar tie gets), that the forces in the tie and rafters are much greater than can be practically designed for.

Good Luck!!

 

[tdsull]

I agree completely with houseguy. The one third point is a good start but often falls short.

ctcry is also dead on...there is a distinct difference between collar ties and rafters ties.They are interchanged to many times in the wrong situation.

Sum moments about the ridge……

 

[RARWOOD]

easleyml

You indicated that the builder has deviated from the requirements of the building code. Based on that statement I am assuming that in addition to the sloping rafters you also have ceiling joists at the rafter bottom spaning from support wall to support wall.

In a system where you have rafters sitting on the wall top plates with ceiling joists running from one wall to the other, collar ties are of dubious value. In the past some model building codes required them in their wood frame conventional construction provisions.

My understanding is that the use of collar ties was based on common framing practice in some parts of the country in the past. As a result that practice was included in some building codes with out a specific engineering reason to do so.

When you have rafters and ceiling joists collar ties generally are of little or no value. As others above have indicated to make the ties work as part of the system requires large connection forces which are probably not achieved through the tie nailing common in the field.

Note if you don't have ceiling joist at the bottom of the rafters then you have rafters with a raised tie member. In that case all the members and connections must be designed to carry the applied loads.

 

[easleyml]

Thanks, your comments are very helpful. By the way we do not have any ceiling joist, so the collar ties will take horizontal thrust loads. Do this change anyone's response now? Thanks

 

[rgerk]

easlyeml,

You need to think long and hard about this design. If you only have rafters and collar ties (up at the ridge) then you have nothing that will resist the thrust load or deflection at the top of the wall. If the contractor doesn't want ceiling joists, the only option you have is to install a structural ridge beam or installing something at the top of the wall that can span corner to corner that will handle the thrust force and associated deflection.

Good Luck. You have a battle of you.

 

[mbullism]

I too always thought the collar ties were to resist ridge seperation either under unbalanced snow loads or wind uplift... I wouldn't omit them without replacing them with steel straps across rafter pairs-

easleyml: I'd be very surprised if omitting the rafter ties (ceiling joists) at the plate met code without the provision of a load bearing live ridge... .02