Stabilising a Spreading Roof

[Flyfisher]

Rebuilding the roof, rebuilding the walls, replacing all the lintels . . . sounds more like a total rebuild than a restoration. As mentioned, if time and money (and enthusiasm) are no object then it's certainly an option, but I find myself agreeing with the previous posts. Clearly the roof needs to be stabilised and made safe but, from my amateur perspective, that should be fairly easily accomplished with suitable metal rods. I would want guidance from a structural engineer of course but from what I've seen of old houses it's a time-honoured method of repair - we have plenty of old iron helping to hold our roof together and it seems to have done a good job for a couple of centuries at least. As for the wonky wall, we have a few of those as well. Again, I'd be seeking professional advice regarding stability and safety but if they pass those tests then I'd leave them as they are as they don't look inconveniently wonky.

This sort of situation highlights how many different ways there are to deal with such 'imperfect' properties and much will depend on the feelings of the owner, but it would be a shame if the character that attracted the owner in the first place ends up being lost as a result of substantial reconstruction.

 

[Feltwell]

Agree with the above. Strengthening in the form of extra timbers, tie bars etc should negate any need to replace the entire roof and certainly any need to rebuild the walls. Done properly the strengthening can be visually OK - patress plates on tie bars can be decorative for instance. The site below sells joist straps as well.

http://www.redgwick.co.uk/Wallplate Catalouge.html

 

[billybuntus]

I'm sure it would provide great satisfaction to pull the roof off and redo it 'properly' but I think steel is your friend in this instance.

Maybe £100 tops to have some appropriate straps made up by your local metal company to your specification. I'm yet to see any building movement that can't be saved by a bit of steel (or lots of it in one instance).

 

[88v8]

And beware of professionals' arse-covering OTTness, especially if you ask them to put anything in writing.
It's only a big shed... you can build a shed, I'm sure.

Ivor

 

[malcolm]

Steel bits (and a little more wood) it is. The roof looks much better now all of the plasterboard has been removed and the structure is visible. There is a little rot at one end where the chimney has cement instead of flashing but otherwise everything looks OK (apart from the issues already discussed).

 

[Penners]

Agree with the above.

Me too. I would never consider stripping the roof unless there were absolutely no alternative. A reinforcement plan will save you a whole heap of money (and time).

 

[CliffordPope]

it must be the rafters have a better connection to the wall plate than the tie beams. Which means they aren't joined together properly.

Exactly - the tie beams aren't forming connected frames with the rafters, but merely sitting uselessly on top of the walls while the rafters push the walls over unopposed.

It seems to me to be missing principal rafters - substantial timbers forming rigid triangles at each tie-beam. Then there should be another purlin much lower down, and the secondary rafters secured to the two purlins. That way it's impossible for the roof to spread.

 

[malcolm]

I've been putting the roof off for ages. I was let down by the structural engineer (he was too busy) and I'm not using the other one. So making it up as I go along.

I've come up with a way to extend the tie beams and connect them to the wall plate using plates each side welded to angle sitting on the wall plate (photo below). I'll add a support for the bottom of the tie beam when I weld it all together.

I've not decided how to attach the angle to the wall plate. My original thought was some more angle on the other side of the wall plate and bolt through. But it's fiddly as there is corbelled brickwork on the other side. The strongest option would be remove the corbelled brickwork, bolt through to angle iron on the other side of the wall plate, fix the angle to the wall, then replace the bricks. Alternatively I could drill straight through the bricks and have the angle on the outside of the bricks. Easiest would be lots of screws on the inside only. I have enough angle for a continuous length on both walls if necessary.

Anyone know how this sort of thing is normally fixed? My rough calcs suggest bolting through would give me a safety factor of about 100 times so perhaps overkill.

 

[Penners]

It's a bit difficult to see some of the fine detail in your photo, but I'm assuming that your reinforcing plates are on each side of the tie beam, and are welded to what looks like a length of angle resting on the wall plate.

If so, then I think you can satisfactorily secure the reinforcing plates to the tie beam by drilling three or four holes in each steel plate and inserting coach screws almost the full width of the beam, from both sides. You could drill matching holes in each of the two plates, and in the beam, and pass bolts right through the whole assembly, but this will make life quite tricky when it comes to alignment (unless you can insert a single drill right through the first plate, the beam and the second plate in one pass).

To fix the angle-iron to the wall plate, the ideal would be to pass coach screws vertically downwards through holes in the horizontal face of the angle into the top of the wall plate. However, I don't know how much this idea would be hampered by the angle of the roof. If it's impossible, then I think you'll have to drill the vertical face of the angle-iron and pass coach screws horizontally through into the wall plate. Obviously these screws will be in tension, relying on the grip of their thread in the wood, but provided you're fairly generous with their number and diameter, they should give a decent hold. In this case, your idea of running the angle the full length of the wall plate - coach-screwed at regular intervals - would seem beneficial.

By the way, that plate cramped to the tie-beam looks like a pretty hefty piece of steel! Given steel's great tensile strength, I think you could safely reduce its thickness somewhat, if you want to save weight.

 

[malcolm]

Once again that sounds more sensible than my original plan. The steel is 5mm but looks thicker in the photo. I can get to the top of the tie beam by removing the lower 3 courses of tiles which will need to be removed anyway to add some sort of eaves ventilation. A single coach screw in shear should be strong enough so a bunch of them ought to do the job.

 

[Penners]

That sounds very promising. However, 5mm plate for the tie beam "shoes" honestly sounds like overkill to me! I reckon 3mm would be ample, and would save weight and money.

 

[88v8]

Would agree that thinner plate would be OK, however its thickness can be exploited.

Fastening to wall.......... why? The rafters are nailed to the wall plate, are they not?
The problem, it seems to me, is the lack of fixings between the rafters and the ties.

So, I'd put that plate on the other side of the tie beam, and bolt through it and through the rafter. Two ½” coach bolts, one near the top of the tie and one in the middle. That thick plate won't deflect enough to worry about, even if it's slightly overhanging the ends of the ties where they've pulled away from the rafter.
Bolting at the top of the tie rather than the bottom will discourage the rafter from splitting, although that's unlikely anyway as the load is oblique to the grain.
You'll then have linked the tie beams to the rafters and created an effective A frame, so no more spread.

I wouldn't worry about additional fastenings to the wall plate at all. However, a couple of coach screws would do no harm, say every other tie.

And if you bolt through the tie beam/plate in three places there's no need for a bottom support btw, it's a tension load.

I wonder if the bodgers who rebuilt this roof are still in business.... makes one wonder what trail of mess they've left behind them.

Ivor

 

[Penners]

Fastening to wall.......... why? The rafters are nailed to the wall plate, are they not?
The problem, it seems to me, is the lack of fixings between the rafters and the ties.

So, I'd put that plate on the other side of the tie beam, and bolt through it and through the rafter.

I'm afraid I can't agree, Ivor. Yes, the rafters should be nailed to the wall plate, but given the dodginess of the rest of the roof structure, we have no way of knowing how effectively this fixing has been made. And by attaching the ties to the rafters you are potentially placing considerable additional stress on that fixing.

On the other hand, attaching the tie beams directly and securely to the wall plate seems to me to be a much safer way to proceed.

 

[malcolm]

The are no principal rafters left, only multiple weedy ones, so I'm aiming to attach the tie beam to the wall plate and strap all the rafters to the wall plate. The wall plate is quite beefy so ought to be up to a structural role but there is a risk of spreading between the tie beams as the wall plate bows. Structural engineer suggested lining the inside of the rafters with ply to take load diagonally from the middle of the roof to the tie beams and take bending load off the wall plate.

Further up the roof I'll add collars between all of the rafters and fit a ceiling.

It is annoying that the work was done so badly. There was a lot of quality work done to the place in 1910, but everything done since then sucks.

 

[malcolm]

How easy is it to replace a wall plate? (the bit of wood the rafters are fixed to) Most of the wall plate was replaced in the '80s but an earlier 4m section remains and that has dry rot as it wasn't ventilated. Would it be a complete roof tile off job to reduce the weight to replace a section of wall plate safely or might roofers be prepared to prop some temporary structure in to hold things in place?

If it is a complete tile off job how far should I go? I was getting comfortable with the rest of the roof and am concerned about restoration spread.

 

[Penners]

A 4m length is quite an undertaking, but I wonder whether you could tackle it in shorter lengths, using half-lap joints to join the individual sections together. You might be able to acrow-prop perhaps two rafters at a time, relieving their pressure on the plate sufficiently to be able to replace that short length... and then repeat the process progressively. The half-lap joints would, of course, best be placed between rafters, rather than immediately under them.

If the existing plate is badly rotted it may be fairly easy for you to break/crumble it away from the fixing nails that presumably come down through the ends of the rafters and into it. You can then cut the nails off. Fixing the rafter to the new plate could be achieved with sturdy angle-plates on each side.

I hasten to add that these are amateur musings, made up on the spur of the moment. You may feel more comfortable about the plan (or any plan that you finally adopt) if you get a nod from a structural engineer.