Engineered Truss designed as raft collar tie setup 1

[Said the Sky]

Hi guys I have a client that requested me to design his trusses (typically I don't do this as I am a base building guy)

so I went through other threads regarding rafter tie design, and was wondering if anybody knows why in SAP2000 if I model the two supports as just rollers (only resisting vertical load and no thrust at the wall and rather connection) it doesn't gives me entirely wrong numbers (magnitudes way off)

however if I change them back to pinned, I get what is shown in the attachment. I guess my question is do you guys design the rafter as full span from wall to ridge beam or does the rafter tie provide some sort of restraint and reduce the rafter span as SAP2000 is doing? if so the bending moments in the rafter beam is significantly reduced.

also how would you guys go about the connection where the king post and rafters meet the ridge beam? technically there is only compression force coming form the rafters, and the king post and the two 45deg angle "braces" are non-structural.

 

[Said the Sky]

more info

 

[Said the Sky]

more info

 

[le99]

If you are not a certified engineer, I would relate the task to a qualified person/company to perform the design. If you are an engineer but lack roof truss design experience, you shall open and review what you've learned in school, then decide what to do.

 

[Said the Sky]

I am a certified engineer, but don't do trusses but understand enough to research and design it and find it sort of interesting to do and add to my skillset. I am just curious if people do use the rafter tie to restrain and reduce the rafter span down which will reduce the bending/deflection allowing the use of smaller members. Essentially I am looking to for experience engineers who has done this before to see what is common practice.

Having said that - I just designed it like I usually do just to be safe, and used the entire span from top of wall to ridge. The rafter where it meets the ridge beam we typically don't spec any special connections there due to only compression forces at the top that cancel out, and collar ties used for unbalanced loading. (in this case we have a metal strap at the ridge)

 

[XR250]

You mention a ridge beam but I don't see that in your model. If you have one, then this is not a truss. If you don't, then the collars are in tension and do not "reduce the span" of the rafters.
Also, you should not have a moment at the peak. I would take le99's advice.

 

[Said the Sky]

Ridge beam is more for looks as this is all exposed and no metal joinery to be exposed either. the span is too large to use his member size as a ridge beam and there is no post at the two ends to support it. Hence I took the approach to use rafter tie, but that ridge beam is blocking my rafters to meeting at the ridge where I could use internal plates and bolts to join then but still hide the connection. But I don't think there needs to be any joinery at the apex since its all compression. or may just ask him to see if he can lower his "ridge beam" a bit.

 

[Enable]

Your tie is so high up in the structure that it probably doesn't do much good as a rafter tie and is more likely to perform as a collar tie (stops truss from separating under uplift but doesnt do jack about spreading under thrust). Maybe it's at the very edge of acceptable (eyeballing it maybe just a hair less than 1/3rd up? looks higher to me though) but I'd be careful with the it.

Also, right now it looks like you've detailed both supports identically but one should be pinned and the other a roller (won't hassle about the practicalities of that as you can find copious threads on it). Here's a great thread on the topic. Here's another link to how to model the truss (picking up on XR250's comment re: moments at ridge. Most would model that with a hinge)

BTW here's a guide for truss design that I have found helpful in the past.