Help with lean to roof spec please

[RonnyRaygun]

Those rafters will be trying to push the wall out. Tying one end of them won't alter this basic principle

What am I missing? This scenario is the same as having a ridge beam (not a ridge board).
The ridge beam prevents the rafters from sagging at the apex. It's the sag at the apex that induces forces in the walls (if the rafters aren't tied).
As the ledger board can't sag, you've effectively got a ridge beam, so no horizontal thrust is present.
To convince myself of this, I've just set up the scenario in QSE. Result? No horizontal forces, just vertical ones...

 

[catlad]

As long as the rafters stay attached to the house wall then I cannot see any outward thrust going on, so do not just rely on skew nailing to your bearer plate

 

[^woody^]

This scenario is the same as having a ridge beam

I'll need to look into this.

Many BCO's I come across wont accept a mono-pitched roof with no ties or no purlin.

It was explained to me years ago that its not just about the support of the top end of the rafter, but also to do with some lateral movement. In addition, something to do with the fixings being OK for vertical shear but not so much for being pulled out gradually.

I can see your point though about a similar concept to a ridge beam

 

[catlad]

For added strength, I put a rapier star fixing through the rafter and bearer plate straight into the brickwork at about 45 deg to the wall, you can get them up to 9" long. As long as the rafters are covered in ply this will link the rafters to the house.

 

[RonnyRaygun]

This scenario is the same as having a ridge beam

I'll need to look into this.

Many BCO's I come across wont accept a mono-pitched roof with no ties or no purlin.

It was explained to me years ago that its not just about the support of the top end of the rafter, but also to do with some lateral movement. In addition, something to do with the fixings being OK for vertical shear but not so much for being pulled out gradually.

I can see your point though about a similar concept to a ridge beam

I've been thinking about this. I guess you have to treat the connection to the ledger board as a pin 150mm or so from the centre of the wall. As such there will be small horizontal forces, although with further QSE checks it seems that they are tiny and the horizontal displacement is virtually zero.

 

[RonnyRaygun]

This book confirms my initial suspicion that no horizontal thrust is present.

 

[tony1851]

A most interesting extract and well-worth reading.
However, it does raise 2 points;

1.The rafter should be firmly fixed at the top and there is often a question on how this should be done. Somehow, just nailing to a pole plate doesn't seem 'right', if only because of the possibility of splitting.

2. Why do truss manufacturers make glide shoes for raised-tie trusses? Almost certainly because the lower part of the rafter (ie below the tie) bends as the weight of the tiles goes on. The implication is that there must be some lateral force (however small) induced by the bend. I know people will chime in saying trusses behave differently to rafters, but the principle of a length of timber deflecting under load, and pushing out, must be the same.

Interestingly, the diagram shows a gap in the birdsmouth of the rafter. One wonders if the model would work with a rafter of longer span (therefore having greater deflection) and with a tight birdsmouth at the top?