Roof tie beam - calling people with structural nous

[Tim Watts]

As many will remember, I'm fixing up a bungalow, circa 1955, very bodged dormer conversion circa 1970s. We're doing the upstairs floor now. The main floor was strengthened with 8x2" joists run to all 3 wall plates (2 external plus centre wall). However, in some "alcoves" this was not done

- the 4x2" ceiling joists were packed up with more 4x2 (badly). This is NOT a cause for concern in itself - ceilings are fine, old ceiling joists are very good quality wood and spans are modest.

Here's the layout in the south east corner (SE corner is top right, black lines are main walls, other structures annotated):

The problem is the tie beam on the right side. These probably ran all the way through originally - but were cut short when the dormer was installed. Clearly not a problem as it's been happy for 40 years.

But we found woodwork had eaten the section marked "Woodwormed" in pink in the diagram above. Looks like that happened a long time ago - no evidence of current activity and all surrounding timbers completely untouched.

So we cut out the worse section leaving the green bit which is 100% sound. There are 3 more such tie beams, the north east one is shown in the diagram on the left in green.

Given it is probably fixed to the rafter with 1-2 rusty 4" nails, clearly it cannot be under *that* much tension, but it seems wise to consider if the missing bit was contributing useful work.

On to the new plan:

In order to strengthen the floor, we've put in a new 4x2 on its side to the right also in green, screwed through to the ceiling rafters. New

4x2" joists will go on top of the ceiling joists, half lapped onto this and glued and screwed to the ceiling joists. This will also form the base plate of a small stud wall.

Obviously not how you'd do it if starting from scratch, but it should be more solid that the original which stood having a 1/4 ton cold water tank (gone) plus the usual junk stored there.

The actual thing I am wondering is how much of that tie beam it is worth putting back. If I add the "Optional noggins" in red as shown, but bolt them through to each other and onto the end of the tie beam with heavy brackets and M10 bolts, I could effectively reinstate the beam as it was last week.

OTOH the new "4x2 on side" is tying the ceiling joists anyway and the tie beam is fixed to the same joist at the end. Not only that, but all rafters are tied back to the first ceiling joist as well as the wall plate to some extent by the vertical yellow 4x2 straps which are forming a horizontal frame which must be pretty stiff in the sideways direction.

I've never been sure how much work tie beams actually do - and the wall plate is pretty substantial (4x3" IIRC).

Anyone dare venture an opinion: a) Belt and braces; b) Nah it's OK?

Cheers,

Tim

[Capitol]

I wonder what it would be like to have a viewable picture?

[Tim Watts]

Are they not working for you? I did check in Chrome Incognito and they seem to be publicly viewable...

[John Rumm]

TBH the tie beams don't do that much... they mainly give lateral bracing that makes it harder for the joints to twist or buckle. They can't give much vertical support since as you say they are only held by nails acting in tension.

You could replace them with noggings or, better still, herringbone (traditional wood or modern metal)

[harry]

Tie beams are in tension. Their purpose is to resist the "spreading" effect generated by the weight of the roof.

The brick outer walls must never be subjected to lateral forces, only vertical ones.

So the tie beams are essential and they must run from the rafter/ wall plate junction at one side of the building to the corresponding junction on the other. Especially important with tiled roof which can weigh tons.

If you're going to bolt timbers together, use a timber connector trapped between the timbers.

This makes a much stronger job. You will need the big square washers under bolt heads and nuts to force the spikes into the timber.

[Tim Watts]

Thanks John,

We'll be adding lots of noggins (and have done where it does not impede plumbing) :)

Possibly a slight misunderstanding - I was wondering how much *tension* tie beams were under in a simple hipped roof?

I know that due to the triangle construction, there is some tendency for the roof to spread outwards and the tie beams countered this.

What I am not sure of is how much restraint is required. I've heard of some traditional roofs that don't have ties, whereas all the ones I've seen (not many) usually have 2 at the 1/4 and 3/4 position (presumably the centre wall plate provides a tie at the 1/2 way point).

Cheers,

Tim

[John Rumm]

That's where herringbone can help - leaves more gaps!

Ah, yup see what you mean...

Again I think the answer is possibly less than you might expect, since many hipped roof places are also semi detached, and hence the ties do not pass through the party wall - they in effect don't actually tie one side of the structure to the other. In yours they could, but again you are talking about a couple of nails in shear on the end of a timber where it itself is also weak in shear.

The wall plate ought to be well nailed down (and these days strapped as well) to stop that moving (hence transferring some of the thrust to a bending moment on the wall plate, and lateral push on the walls. The heavy bracing work is normally done by the purlins and any A frame style ties between the rafters (which in your case are probably holding up the upstairs ceilings!)

They are normally placed mid span of the ceiling joists (or sometimes

1/3 and 2/3 if there are a pair front and back (i.e. 4 total) on larger roomed properties.

[Tim Watts]

That's reassuring - thanks John.

We are improving the purlin supports too (cut free end, the other is well nailed to the hip beam).

It is quite nice taking out a certain amount of the shonky crap from the

70s and putting back something that is a) straight; b) might actually do something useful!

Cheers :)

Tim

[John Rumm]

While this is all true, I would suggest its also not what Tim is discussing here...

With a hipped roof, you normally have the ridge beam parallel to the spine wall, and that will usually (but not always) run along the centre of the longer axis of the building. The floor/ceiling joists themselves run perpendicular to the ridge, and form a strong tie right from one side of the building to the other by tying the base of every rafter pair together. This creates two opposing pitched roof faces that are acting as a self supporting structure, with all thrust directed straight down.

The addition of hipped roof sections onto the side of this structure typically imposes additional mainly downward loading onto the ridge, the hip rafters, and the wall.

I believe Tim is describing the individual small timber that is run across the top of the main axis floor/ceiling joists. Its main purpose in life is giving lateral restraint to the ceiling joists. A pair of

4x2" (if that) alone are not going to provide any significant lateral support to the whole roof structure, especially since they are only nailed to the wall plate and, if you are lucky, one rafter.

[newshound]

Be careful, the strength of the joint is *not* that of a couple of nails in shear, it is the friction of two large areas of wood held together by the *tension* in the nails. And when these get rusty, they lock in the original tension very well.

[Tim Watts]

Well - I reckon there's probably not much to lose by sticking the red noggins in (2nd drawing) but bolting them together so they are strong in tension and continue the tie beam. At least I'll get back to how it's been for 40 years (well bit better as the woodworm bit was like a sponge for about 3-4 ft).

Having said that, I was wrong when I said "3 other tie beams" - I did have a look over the NW corner where the stairs are - bugger all tie beam there (because the stairwell chopped it out) and that's been like it for years and the roof's not done anything weird.

[Tim Watts]

Was going to say:

look just the part - long side down the noggin/tiebeam to move the bolt hole away from the end, short side to the rafters...

[harry]

No mention of hipped roof. Methods of construction vary with hipped roof.

[F Murtz]

He might mean actual picture of the beams.

[Tim Watts]

Happy to upload if Capitol confirms (because I'll have to spend some time digging for the ones I took a few weeks ago).

I'm actually going to run this lot through Superbeam to see what I reckon that section of floor is good for.

I am excess of 100kg and I can stand on a single ceiling joist mid span without excessive deflection (well the ceiling did not crack!) - and I have done may times.

So the crudest estimation is, if all ceiling joists are tied together with noggins at both ends of the floor area, it will probably be good for 800+kg - that's more than you could get lardbutts like me in that space due to the rapidly diminishing ceiling height, the highest bit of the ceiling being over the centre load bearing wall too.

Cannot think why the original willoughby didn't just run a full set of

8x2's right through. Actually if it were me and engineering calcs permitting, I might have just been inclined to run an RSJ or flitch beam perpendicular *underneath* the joists mid span into the rooms below - got extra head height to boot.

[John Rumm]

Well, Tim's diagram shows rafters on both sides of a corner.

(and if were a gable roof, then the situation would be even better since the tie beam in question is running perpendicular to the ceiling joints)

Indeed. Just as well I have walked round Tim's place and looked at it in person ;-)