Siccors Truss Connection 2

With A Siccors truss being design with one end as pinned and the other as a roller to eliminate the horizontal force onto the wall, how is the connection made to the wall as not to put the horizonal force into the wall. I understand that a slip joint can be made from the truss to the wall on one end of the truss. Doesn't this allow movment at the top of the wall due to wind forces? What type connection should be made. Are there any example or illustration on the internet.

Thanks

Jason

 

[ishvaaag]

Rollers, plates on teflon, rubber bearing plates layered or not on pots or not etc can be used. Some -loose- clamps above what slips on the bearing plate going upwards if such was the case...in many cases but not all can prevent this.

It is also common to enter some minimum lateral force even at the gliding support, especially if friction, since some always exist.

 

[JAE]

If this is a wood framed wall and wood truss, the wall will most likely not provide a whole lot of lateral restraint and thus, the truss will act as though it is pinned. You probably have to connect the trusses to the wall for stability against lateral wind forces on the wall, or against some minimum lateral pressure on interior walls (the UBC requires 5 psf).

With the actual fixed connection (and idealized pinned connection), you can calculate what the lateral deflection is at the top of the wall. You need to make sure that the wall can take this distortion without undue distress in finish surfaces, etc.

I've never seen teflon, rollers, etc. used on this kind of construction. If you have a more significant structure, such as a concrete or steel building, then, if you do use some kind of detailed slip connection, the wall itself must be laterally stable on its own. It (the wall) cannot be designed as though the truss/roof is bracing it, but rather, designed as a cantilevered wall or a wall with some other type of independant bracing.

 

[ishvaaag]

JAE must be right, I don't completely master the english language and I was speaking in general way.

 

[phuduhudu]

This is a common misconception which arises out of applying models in isolation. If you look at the structure as a whole your walls will probably not restrain the truss against lateral movement so you do effectively have a sliding support as you modeled even if it does not slide over the wall plate. If your wall is so rigid that it provided restraint to your truss then you will probably not need to restrain the top of the wall since the top of the wall is the thing doing the restraining anyway.

Of course there is an interesting exception in that trusses over or near end walls or cross walls are likely to be totally restrained and should not be modelled as having one sliding support. This is not likely to be a problem for the wall but you need to check that the truss can take the different set of forces. You also need to check that the difference in deflection of adjacent trusses with different support conditions is acceptable.
Carl Bauer

http://www.bauerconsultbotswana.com

 

[steve111]

This is an interesting problem left to the Engineer of Record when specifying "'Pre-Engineered' wood roof truss, submit sealed drawings for review prior to fabrication" etc., on your projects. It goes to the heart also of whats wrong with the Truss Plate Institute members, those that manufacture truss plates and sell plates and use of their truss software to manufacturers of trusses, and Manufacturers of trusses. One should really take a close look at the disclaimers that accompany a set of the sealed shop drawings that come with the trusses.

The truss is designed as an individual structural element for the loads specified by 'others', with the requirement for review of the drawings and field execution by 'others'. The pinned end, roller end is 'economically' a convenient model for them, with the only TPIC requirement being to limit the roller end horizontal deflection to 1". Pinned, pinned would cause their software to abort the job, making dimensional lumber and their plates unmarketable for this type of application.

What's the jobsite reality of constructing a pinned, roller connection? None that we have found so far. We did find one cold formed steel connector that was designed to allow horizontal movement but issues of corrosion between the sliding plates, friction, and uplift seemed to be lost on the manufacturer. Issues of serviceability limit states, ie deflections and finishes, seem to be lost on the truss plate industry. Using pinned, pinned connections of course negates any liability of the truss plate supplier and their manufacturer as you have completely changed the loading on their truss design. It is truly a convienent model on their part.

The Truss Plate industry really needs to develope 3 dimensional models that account for roof diaphragm action, load sharing systems, bearing connections and wall (both exterior and interior) stiffness and interaction, etc. I suspect that the BELIEF of many is in the rundundancy of structural wood members at two foot or less centers (read Ultimate Limit States) collapse won't occur, and that between normal shrinkage of wood members and cracking of finishes (read Serviceability Limit States), nothing a little 'mud' and paint won't fix, and the Truss Plate Industries knowledge that most of their designs are never reviewed by anybody remotely qualified as required by their shop drawings, that this is all 'someone elses' problem (read Engineer of Record).

It's a little early in the morning here, and re-reading this 'rant' one can tell i've had a few to many coffees. I would be interested in hearing from others with similiar experiences and 'solutions' they have found to the scissor truss pinned/roller connection.

 

[JAE]

If the truss designer assumes a pinned end, that is usually conservative as any resistance by the wall to lateral movement would decrease tension in the bottom chord and most of the connections. The top chord would see more compression but is usually quite well braced laterally.

Steve111 - while I agree with you about modeling trusses to match reality, I don't think you can "turn" this industry any time soon. Most of those "designing" trusses are not engineers and those that actually do the sealing of truss designs cover a large sales area. It would just be way too complex for this industry to respond to a 3D system of designing.

ishvaaag - no need to apologize for your English....I work with only one language and I always respect those who can learn more than one, like yourself.

 

[CTruax]

Steve111(Structural),
Your opinion is all deflected about truss + pin + design.

Scissor (and all other configurations for that matter) are typically designed Pinned-HzRoller because that yields a more conservative design. If you were to take a 38' span scissor and analyze it Pinned-Pinned at the reaction points you would notice a significant decrease in material requirements: Say 2x6 chords --> 2x4 chords, and significant metal gusset plate (steel) savings. Truss fabricators would love to build your EOR projects that specifiy Pin-Pin and let you handle the horizontal thrust load at the walls.

Fact from friction be that a quality truss fabricator will run the design internally as multiple load cases and supply a component that meets both reaction conditions. An EOR such as yourself might even REVIEW the truss calculation and potentially request such a design criteria if absent.

I would respond to the initial post with three categorical cases;
1. Design snow load = 385 psf (such as Tahoe)
2. Design Live load = 16 psf to 25 psf (90% of US)
3. Conventional framing - Interior support - Steel truss

1. High snow load area engineers expect trusses to 'breathe' with seasonal changes. Usually spans are shorter and interior bearings more frequent with OC spacings of 16" and 12" and 2x8 chords common. The increased truss mass still responds to the design limits a truss engineer may specify (say 1" Hz max) as in case 2, but the expected movements do indeed happen with each winter storm. Special attention is needed for this expected movement.

2. In lighter live load areas this problem occurs much less frequently. Most vertical and Hz deflections occur during the first 7 days of construction due to dead weight materials being applied. Truss design engineers typically flag designs with special Hz movement notes in an effort to get the EOR to think. I don't know of a TPI supplier that would not supply an EOR additional load/reaction cases gratis.

3. Limit the potential for deflection problems by practicing the following. If possible
-Make the truss-plate attachment with dead load in place.
-Use a horizontal-slotted framing clip at the attachment.
-Make sure slot-nail is placed to allow anticipated movement
-Specify flexible blocking between trusses.
-Install ceiling rock held back 1/2" and wall rock flexible
-Design double top plates for horizontal thrust.
-Design home with limited truss run between collar-tie walls
-Add a beam at scissor apex and specify two mono-scissors.
-Expect walls to plumb-out at each truss heel
-Specify strongbacks at start and end of truss run
-Install a metered heat-strip on the bottom chord so as to reduce moisture content and increase shrinkage in proportion to deflecton during heavy snows (pat-pending...

 

[phuduhudu]

steve 111, I think your words for the truss industry are a little harsh. I am a consulting engineer myself and I would say it is us who are at fault if there is any lack of oversight in the whole design. Why ask the roof truss manufacturer to produce 3D models of the whole building? Surely that is our job. There should be one structural engineer with overall responsibility for a building and he should either produce the truss design himself or fully review the manufacturer's design.

Carl Bauer

http://www.bauerconsultbotswana.com

 

[phuduhudu]

steve 111, I think your words for the truss industry are a little harsh. I am a consulting engineer myself and I would say it is us who are at fault if there is any lack of oversight in the whole design. Why ask the roof truss manufacturer to produce 3D models of the whole building? Surely that is our job. There should be one structural engineer with overall responsibility for a building and he should either produce the truss design himself or fully review the manufacturer's design.

On the issue of sliding details, we have to be careful not to let our models steal our common sense away. We do model one end of a truss as a sliding restraint normally because the wall cannot provide the required lateral restraint for it to be a pin. I have seen some very unconvincing attempts to produce a sliding restraint dreamed up by an engineer who is thinking of text book diagrams and not reality.

Carl Bauer

http://www.bauerconsultbotswana.com

 

[steve111]

ctraux, hello.
well yes we do review the shop drawings as submitted. with regards to your example one to the initial query in this thread ...i believe it is this ' special attention is needed for this expected movement' that is of interest to myself anyways.
deflection in light wood framed structures is purely a serviceability live load (snow and/or wind) driven calculation. i will dig through my project notes we were involved in several agricultural facilities with 60' and 70' clear span scissor trusses @ 4' o.c. DL 12psf., LL(snow + rain) 46psf. the buildings were post/beam framing with wood posts @ 16' centers and open side walls. will look for the truss manufacturers drawings but a believe the hz reactions per truss were in the neighbourhood of 2500 lbs (specified loads, unfactored reaction).
Not sure what your metered heat strip is all about but we had a winter ice storm here several years ago in which the electricity was off in a significant geographic area for between 1 and 3 weeks. significant number of structural collapses, significant amount of damage deflection related, and piles of people unloading their roofs with shovels (fairly low tech but very effective). calculated roof loads about 80 to 90% of design load. Oddly enough as an aside to this all the roofs (cold formed steel sheet metal roof cladding)had been designed as "slippery roofs" with a slope factor of less than 1. we had a combination of rain and snow over three days and the whole load (snow, rain and ice) was frozen (stuck) to the heads of the steel screws holding the roof cladding to the purlins. not very "slippery" as it turned out.
carlbauer: yes agreed, a little harsh for sure.

 

[CTruax]

Explanation from Alpine's site

http://www.alpeng.com/Publications/GoodCon/Summer_1997/gcsu9704/gcsu9704.htm

 

One should never confuse the model used with the actual construction.

Building practice is almost pinned and almost pinned.

Pinned and roller as a model is conservative relative to building practices.

Pinned and pinned as a model is dangerous relative to building practices - it is impossible to build pinned and pinned with wood construction.

 

[surfsail]

assuming light wood gage steel trusses ~24" o.c. if you are in an area with wind loads, providing a slip connection at bearing ends may not be a reasonable for designing walls or bld'g diaphragm connections to roof...
conservative to (i.e.) design truss for both 'pin/pin-slide' and 'pin/pin' conditions. 'slide' governs top & bottom chord design.
check location of walls intersecting => rigid spots
dont forget to get the architect to detail for stress and lat'l movement if there are beams subjected to weak axis bending with glass or similar walls below. diaphragm action helps. quantify it w/ defl compatibility/stiffness....
you need to look at all of your conditions & design/detail accordingly..