ICF & Sealed Attic - Need Help Fast
Last Post 14 May 2009 05:59 AM by Buntly. 16 Replies.
Printer Friendly
Sort:
PrevPrev NextNext
You are not authorized to post a reply.
Author Messages
mlennoxUser is Offline
New Member
New Member
Send Private Message
Posts:66

--
12 May 2009 07:02 AM
I'm going down to city hall tomorrow to apply for a building permit and I'd like to get a more accurate cross-section diagram of the top of ICF wall - roof truss assembly.  We're doing a sealed attic space with lofted ceiling (parallel chord trusses) and spray foam against the underside of the roof deck, similar to Figure 10 on page 16 of this document.

http://www.buildingscience.com/documents/digests/bsd-102-understanding-attic-ventilation

I'd like to do the rigid foam backer as shown, but have it sitting on top of the ICF wall (presumably on top of the outside foam layer, then spray foam as shown.

If anyone has a cross-section of what they've done similar to this is would be greatly appreciated.


dmaceldUser is Offline
Veteran Member
Veteran Member
Send Private Message
Posts:1465
Avatar

--
12 May 2009 09:34 AM
For what it's worth here's the detail from the drawings I did for my house. I added sill seal foam under the sill plate. I'd use OSB for the backdam. Stronger, easier to fasten securely, and cheaper. The extra foam won't add enough R value to make it worth using it in the eave, IMO.


Attachment: Wall to roof connection.jpg

Even a retired engineer can build a house successfully w/ GBT help!
mlennoxUser is Offline
New Member
New Member
Send Private Message
Posts:66

--
12 May 2009 11:58 AM
Thanks for this, very nice approach.

I agree the rigid foam is probably overkill.  What did you roof trusses look like in side profile?  I think that with mine I will have a problem with your approach since the trusses extend out to create the soffits.  I believe the trusses will have vertical members over the sill plates that I could mount OSB sheathing against as the spray foam backing though.


TerryJUser is Offline
New Member
New Member
Send Private Message
Posts:40

--
12 May 2009 12:28 PM
Posted By mlennox on 05/12/2009 11:58 AM
Thanks for this, very nice approach.

I agree the rigid foam is probably overkill.  What did you roof trusses look like in side profile?  I think that with mine I will have a problem with your approach since the trusses extend out to create the soffits.  I believe the trusses will have vertical members over the sill plates that I could mount OSB sheathing against as the spray foam backing though.

Probably the heel of your truss is raised to allow for more insulation. These are also call "energy trusses".



Attachment: heel truss.jpg

AltonUser is Offline
Veteran Member
Veteran Member
Send Private Message
Posts:2164

--
12 May 2009 01:45 PM
Since the 1970's I have used the energy truss.  In some cases the bottom chord of the roof truss formed the attachment for the soffit.  Some framers say that extending the lower chord can result in a soffit that is not level since the trusses are not always free from warping and twisting.  Wes, or anyone else, what has been your experience when the bottom chord of the truss extends to make the soffit?


Residential Designer &
Construction Technology Consultant -- E-mail: Alton at Auburn dot Edu Use email format with @ and period .
334 826-3979
wesUser is Offline
Advanced Member
Advanced Member
Send Private Message
Posts:810

--
12 May 2009 05:16 PM
Sorry Alton, I have never used an energy truss built to your description. I have always used a truss were the bottom chord stopped at the line of the exterior wall. The raised heel was accomplished with blocking, and only the top chord extended beyond the exterior wall. I can see the framers' point about the movement of the bottom chord. For some reason the bottom chords do seem to move more than the top chords.


Wes Shelby<br>Design Systems Group<br>Murray KY<br>[email protected]
dmaceldUser is Offline
Veteran Member
Veteran Member
Send Private Message
Posts:1465
Avatar

--
12 May 2009 09:57 PM
Posted By mlennox on 05/12/2009 11:58 AM
I think that with mine I will have a problem with your approach since the trusses extend out to create the soffits.  I believe the trusses will have vertical members over the sill plates that I could mount OSB sheathing against as the spray foam backing though.
The trusses over my living room extend an extra 24" over the wall to form a 40" overhang. The bottom chord though was not used for the soffit. There is a vertical chord at the wall line like you describe. If I remember correctly, what we did there was add a nailer 2 x 4 to the vertical chord and fit OSB between each pair of trusses, one end nailed to the chord and the other to the added nailer. A tight fit isn't necessary as the foam will fill up a small gap. Where the gap was on the order of 1/4" to 1/2" the insulator used can foam to close them up before spraying the Icynene.




Even a retired engineer can build a house successfully w/ GBT help!
dmaceldUser is Offline
Veteran Member
Veteran Member
Send Private Message
Posts:1465
Avatar

--
12 May 2009 10:28 PM
Here's a portion of the truss company drawing for the living room truss.


Attachment: Living rm truss 3.jpg

Even a retired engineer can build a house successfully w/ GBT help!
wesUser is Offline
Advanced Member
Advanced Member
Send Private Message
Posts:810

--
13 May 2009 06:35 AM
dmaceld,
OK, I'm sure not a truss designer, but that truss sure looks like it would put a tremendous amount of outward pressure on the left hand wall.
Did you have to do anything to reinforce that wall and/or have you noticed any outward movement of the wall?


Wes Shelby<br>Design Systems Group<br>Murray KY<br>[email protected]
dmaceldUser is Offline
Veteran Member
Veteran Member
Send Private Message
Posts:1465
Avatar

--
13 May 2009 09:37 AM
Posted By wes on 05/13/2009 6:35 AM
dmaceld,
OK, I'm sure not a truss designer, but that truss sure looks like it would put a tremendous amount of outward pressure on the left hand wall.
Did you have to do anything to reinforce that wall and/or have you noticed any outward movement of the wall?

Outward movement? With an ICF wall! You got to be kiddding!!!!!

The trusses were designed and built by a local truss company and stamped by their engineer. All such trusses are designed to transmit only vertical load to the support points. All lateral stresses are contained within the truss.




Even a retired engineer can build a house successfully w/ GBT help!
wesUser is Offline
Advanced Member
Advanced Member
Send Private Message
Posts:810

--
13 May 2009 01:30 PM
dmaceld,
sorry, I forgot you were building with ICFs.
However, I will argue with you about the lateral loads on the truss. Your statement only applies to a truss with a flat bottom chord. Any truss with a vaulted bottom chord, such as yours, will apply some lateral stress to the walls to which they are attached.


Wes Shelby<br>Design Systems Group<br>Murray KY<br>[email protected]
James EggertUser is Offline
Basic Member
Basic Member
Send Private Message
Posts:411

--
13 May 2009 06:03 PM
True, BUT the attachment at the bearing points transfers the load vertically.

I have never seen a frame wall with scissor trusses requiring any additional top plates to handle outward thrust....it's not there!


Take Care<br>Jim<br><br>Design/Build/Consulting<br>"Not So Big" Design Proponent
BuntlyUser is Offline
Basic Member
Basic Member
Send Private Message
Posts:162

--
13 May 2009 07:36 PM

99.9% of all trusses are modeled with one of the bearings as pinned and the other as a vertical roller. The dwg shows two number R=1148 and R=1350. These are the vertical loads. The numbers U=184 and 189 are the uplift loads due to wind. If there was a horizontal thrust, the truss dwgs would show an Rh number (horizontal thrust). There are very few instances in which a truss need to have both walls modeled as pinned. In the case of both brgs being analyzed as pinned, if the truss is going to deflect downward, the wall must move outward. If the wall is designed to carry the horizontal thrust, it is impossible for the truss to deflect vertically. (kinda like standing a pair of scissors up on the table). I remember one job I did a few years ago had a 24' scissor truss. Both the roof and clg pitch were 12/12. The truss had a 2' heel. The only way this truss would work was if both walls were modeled as pinned. Luckily this was an icf wall. I had to have arxx engineers very that the wall could take 700plf of outward thrust. Would've been impossible with stick frame walls.  All that being said, the reality is that everything moves. It is up to the bldg designer to determine whether the horizontal movement is acceptable. I can't remember, but I think .75" deflection due to live loads and 1.25" total load deflection is considered acceptable in most cases.

Bunt



Bunt
BuntlyUser is Offline
Basic Member
Basic Member
Send Private Message
Posts:162

--
13 May 2009 07:44 PM

Just found this from this from the software company which was used to design dmacs trusses.


Considerations in Specifying a Connection for Horizontal...

Considerations in Specifying a Connection for Horizontal Movement

By Charles C. Hoover, Jr., P.E.



Introduction

There is always the possibility that a solution to a perceived problem brings with it the roots of another problem. The scissors truss clip is one of these solutions. In this short article, the intent is to review the factors relating to the horizontal movement of a structural element, and to review considerations in the design of the overall building for this movement.

 


 

Assumptions of Truss Design

Trusses are typically designed for a bearing condition that has a pin type bearing on one end, and a roller bearing on the opposite end. This means that the truss is able to move laterally in response to the load it is supporting. This movement is due to the flexibility of the structure's construction and is a normal response in most residential and light commercial construction.

All trusses exhibit horizontal movement at bearings. Scissors trusses produce horizontal movements that vary with span, pitch and loading. Most scissors truss designs include an advisory note about the horizontal movement of the truss. This information is intended to assist the building designer in the evaluation of the overall structure.

Assumptions of Wall Design and Construction

The wall supporting the roof or floor system is designed to resist several loading conditions.

• The side wall supporting the truss is designed to resist lateral load applied to the end wall from wind or earthquakes.

• The wall is designed to resist downward loads from material, construction, or snow.

 • The wall is designed to resist uplift loads such as wind loads.

• The wall is designed to resist the lateral loads from such sources as snow drifting or wind.

 

For lateral loads, the wall is designed on the assumption that it functions as a simply supported beam, with a bearing at each end of the wall. The bottom of the wall is supported by the foundation (or floor diaphragm if the building is a multi-story) and the top of the wall is supported by the roof diaphragm.

The reaction at the bottom of the wall is transferred to the foundation through mechanical connections such as anchor bolts or light gauge steel connectors. The reaction from the top of the wall is transferred through the truss to the roof diaphragm.

Conflict of Design Assumptions in the Use of a Slotted Truss to Wall Connector

Several proprietary products have been developed and promoted to provide for the horizontal movement of scissors trusses on walls. The horizontal movement is provided by slots in the connector, into which a nail is placed. The nail is intended to be capable of resisting uplift loads, and free to move in the slot of the connector when lateral movement of the truss occurs. The slot in the connector is intended to allow the truss to move, while the wall remains in it's original position.

The conflict of the movement provided by the slot occurs when the building is subjected to wind load. In this load condition, the truss may be in uplift, and the wall is loaded laterally. During this condition, the reaction of the top of the wall would not be transferred to the truss until the wall moved, and the nail in the slot of the connector contacted the metal at the end of the slot. The amount of this movement would depend on the nail location, and the length of the slot. Also, it would be an abrupt, rather than a gradual movement.

The amount of horizontal movement of the wall under this wind load condition may be unacceptable to the building designer or building owner.

Realities of the Building Design and the Structure's Response to Loads

There is nothing that we can do to escape the fact that trusses (or any other structural element) will move horizontally under load, nor that the wind blows and will place lateral loads onto the wall.

Almost all wall construction is based on the wall being supported by the foundation and a roof or floor diaphragm. In order for the wall to be supported by either means, a solid connection must be provided between the wall and the supporting element.

Based on the building designers' experience with the performance of the completed structure, they may prefer to allow the wall and truss to be positively connected. This is in recognition of the flexibility of the construction of the building, and that a certain amount of movement is acceptable.

Historical Performance of Walls and Scissors Trusses

Design considerations that engineers and building designers use are based on the actual performance of materials and connections in response to imposed loads.

The performance of the walls of buildings supporting scissors trusses has been excellent. In most cases, little if any special considerations for connections between wall and truss have been provided for the lateral movement of the wall due to the movement of the truss. This performance is probably due to a combination of factors: Design loads may be higher than the actual loads imposed on the trusses; the roof and ceiling diaphragms provide considerable resistance to lateral deflection; the resistance of the diaphragm is not considered in the deflection calculations of the truss; and the assumption in the truss design that the one end is a roller bearing predicts more movement than will actually occur.

Recommendation

The overall movement of the truss and the wall must be a consideration in the design of the building. If the building is sited in a high wind region, the load from the wall must be transferred into the truss and then into the roof diaphragm. A connector that is designed to allow lateral movement between the truss and the wall would not be appropriate in this application.

 

Bunt



Bunt
BuntlyUser is Offline
Basic Member
Basic Member
Send Private Message
Posts:162

--
13 May 2009 08:07 PM

I just happen to have the software which designed dmacs truss. For the heck of it I ran his truss in two scenarios, one with unpinned brg and the other pinned. Notice the Rh=1080# on the pinned version. If the truss were analyzed that way, the wall would need to resist about 540 plf (trusses 2' o.c.)

 

Bunt


Attachment: unpinned truss.pdf
Attachment: pinned truss.pdf

Bunt
dmaceldUser is Offline
Veteran Member
Veteran Member
Send Private Message
Posts:1465
Avatar

--
14 May 2009 01:02 AM
Posted By wes on 05/13/2009 1:30 PM
dmaceld,
sorry, I forgot you were building with ICFs.
However, I will argue with you about the lateral loads on the truss. Your statement only applies to a truss with a flat bottom chord. Any truss with a vaulted bottom chord, such as yours, will apply some lateral stress to the walls to which they are attached.

I'll concede, somewhat! If the truss ends are solidly constrained then yes there would be movement in the walls as the load on the truss varies. But in reality that movement is very small, otherwise you'd see all kinds of cracks in the sheet rock. (Maybe that's why crown molding is popular, hides cracks in wall to ceiling joint. ) Most likely what happens is that by the time the sheetrock gets taped the majority of movement will have occurred due to all the dead weight that has been attached, i.e., sheathing, shingles, sheetrock, and insulation. At that point the system will be static. The only flexure thereafter will be from live load like snow. This truss was designed for no more than L/480 live load deflection. If I did the trig correctly at 26' that would bring each end in only about .0005". I wouldn't be surprised if that isn't less than the movement from thermal expansion/contraction of the roof and ceiling from seasonal temperature changes. That's a calculation I'll let someone else do! With that minor amount of movement I'd be willing to bet that any commonly used attachment system will give enough that in effect the trusses can be considered to be on rollers at both ends!

Now think about a flat bottom truss. As it's loaded it too will flex downward, pulling the walls together! That means there is a lateral load at the ends of all trusses, regardless of configuration, but that lateral load exists only when the truss flexes up or down after it is fastened to the walls.

I guess what we tend to overlook is that any house structure will move and twist and turn, but it's usually so little we never see evidence of it.



Even a retired engineer can build a house successfully w/ GBT help!
BuntlyUser is Offline
Basic Member
Basic Member
Send Private Message
Posts:162

--
14 May 2009 05:59 AM
Right on!

Bunt


Bunt
You are not authorized to post a reply.

Active Forums 4.1
Membership Membership: Latest New User Latest: croccohvacusa New Today New Today: 0 New Yesterday New Yesterday: 0 User Count Overall: 35027
People Online People Online: Visitors Visitors: 893 Members Members: 0 Total Total: 893
Copyright 2011 by BuildCentral, Inc.   Terms Of Use  Privacy Statement