Code Experts....

Table 802.5.1 specifies a Rafter Span Adjustment Factor in Note a. 

Hc
---
Hr     Factor
1/3   0.67
1/4   0.76
1/5   0.83
1/6   0.90

The upper limit shown is 1/3 the way up, above the wall plate.  There are no restrictions that say you cannot place the rafter tie above 1/3.  But could you?  The 0.17 to 0.33 range is a Fibonacci Sequence (3,5,8....) that can be extended to:

802                  802   
Frac   Decimal   Fib   Factor      Fac Incr
1/2   0.46            13   0.60       0.07
1/3   0.33              8   0.67          0.07
1/4   0.25              5   0.76          0.07
1/5   0.20              3   0.83          0.07
1/6   0.17              2   0.9   

We know that using the span adjustment factor reduces the allowable span because you are multiplying the table spans by the decimal factor.
If I wanted to place the rafter tie 1/3rd the way up, I would have to find a appropriate span, species, spacing, thickness. etc. using the factor.

But, If I wanted to go above 1/3rd, could I use Fibonacci calculations to further reduce the span and be in compliance with the code?
I would not be interested in 1/2, but I might be in the area between 1/3 and 1/2. 

Thanks, Dave 

I was looking at that sequence for the drive on a trebuchet last week, funny how often that shows up. Short answer... not without an engineer's blessing. You can use the WFCM to go up to 1/2 height though. I posted a link a month or two ago but can't lay my hands on it, maybe someone bookmarked it and will chime in. The heeljoint connection becomes critical as you rise.

(funny how often that shows up)
Fibonacci sequence is a natural progression (found in nature).  Large groups of people will naturally act/behave in accordance with the Fibonacci sequence.  Also the Fibonacci sequence can be seen in the branching in trees, sprouts of a pineapple, arrangement of a pine cone's bracts, etc.  I use Fibonacci retracement in stock trading. 

Found the WFCM online.  I'll see what it says. 

It's only marginally more explicit in IRC 2015 Figure 802.5.1 where it says Hc/Hr = 1/3 max.  That's the only place I saw that says you can't extend the sequence.

Quote from: ChugiakTinkerer on November 25, 2016, 10:29:27 PM
It's only marginally more explicit in IRC 2015 Figure 802.5.1 where it says Hc/Hr = 1/3 max.  That's the only place I saw that says you can't extend the sequence.

Chug,
What page does it say "max"? 
Dave

Fig R802.5.1, right side of the drawing.

My 2000 codebook has a factor of .58 at 1/2 height and .50 for 2/3 or greater and the drawing in fig 802.5.1 is different, this section of code has tightened up.

I see the "max" in 2015 IRC.  I was looking at 2009 IRC.  Hampshire county says meets 2009 IRC/IBC.  Don't know why they don't specify 2015 unless they just haven't updated the web site.   

I kind of agree with the 1/3 max. 

A friend of mine and I were talking about this and other methods to prevent wall spread so that the design for a vaulted ceiling had more latitude.  I told him I thought about cable.  He said he remembered a building on Paris Island that had cable every 8 feet with turnbuckles.   

It is possible that the website is out of date, but more likely that is the version of the code that your state has currently adopted. I'm in VA and we are currently using the 2012 version with Virginia amendments and changes. The 2015 is the current version of the national "model" code. Each state then looks it over, makes any changes they need to and then the state legislature votes to adopt that version into law. My state usually takes 2 or 3 years to do that work before we adopt the latest code. The model code comes out every 3 years. I also agree with the changes in that section.

It is a tension load so steel rod, cable, chain, etc are fine if used properly. You've stepped outside of prescriptive code so technically an engineer is required. I sit down with the building official and talk through where I'm straying from code and explain what I plan to do. It is then his decision on whether he wants professional design.