I am designing a timber roof with canted oak purlins and have designed the roof frame 3-dimensionally.
My analysis assumes that the rafters are connected to the purlins via timber pegs. The analysis concludes that the purlins will resist a vertical and horizontal reaction from the rafters and hence this requires the purlins to be tied at their supports to resist the horizontal reaction. I propose to introduce a horizontal tie member between the two purlins and wonder if anyone has a more sympathetic approach to this problem?

Also, the timber pegs will tend to resist the axial load in the rafter in order not to present thrust at wall plate level. This axial load is 1.0 kN and I am struggeling to justify a timber peg, any ideas?

Hi Clive,

Sounds like you are in a pickle !

There are two ways that a traditional English rafter / purlin roof works.

First is if you have continuous rafters that run peak to plate and pass over the purlin generally with a simple square peg connection at the purlin and generally seated but unpegged at the plate.

Second is where the rafters and purlin are both faced to the exterior of the roof plane i.e. the rafters are interrupted by the purlin and hence require to be jointed top and bottom into same.

In the first case the rafters are supposed to push against the wall plate and not the purlin hence you can discount the purlin peg connector in your analysis. The common rafters in this case will push against the purlin to a greater or lesser extent depending on their relative thickness and position in the rafter cant i.e. when in the middle they will tend to push less becuase of the inherent deflection in the purlin however they will push more when they are closer to a principal rafter. The square pegs are used more for simple location and anti racking rather than accepting and directing roof dead / live loads. These joints will come apart in long term service but this is not a cause for concern.

In the second case only the underside purlin mortice is pegged with the top joint being morticed but left unpegged. This type of roof does need a bit of lateral support and it is not uncommon to see retrofitted birdsmouthed collars inserted across the span of the purlins to help reduce the sag on the purlins.

Either way none of these considerations are building killers. Pegs in these positions rarely break.

I am located just south of the M4 in North Hants (Bramley) so if you need further help just email me or call.

Regards

Ken Hume P.Eng C.Eng.

Hey guys check the Ferry Farm report(s)on the guild homepage. This roof is a heavy weight version of the structure in Kenmore House, Fredericksburg, VA. The Kenmore roof dates to 1765 or thereabouts and is still sound. If anything the orginal wood shakes have been replaced with slate adding tons of dead weight.
The roof structure still stands.

Many thanks chaps.

For such a small axial load, would a peggged half-lap or open mortice & tenon at the the peak solve your problem? Here in the U.S. I would use a timberlok screw (www.olyfast.com) to handle the small load at the rafter to purlin joint.