vaulting a cieling?

[middi]

To simplify matters you could nail another lot of plasterboard over the existing and then replaster?.

I also notice ( if I interpreted it correctly from the pictures) that the older part of the rafter ceiling is also a later addition, the "scantly built" collars have been "spiked" (nailed) against the face of the rafters. This was only common in the 18th 19th century. As you mentioned that the wall plates and end collars are of substantial construction one has to think that the building was converted from it's original intended use ( perhaps no means of heat or smoke = no soot), perhaps only one or two central collar dovetailed clasped and pegged to the pair of rafters in the centre of the room above the rafter purling were fitted . Look for evidence.

Middi

 

[Penners]

ive been thinking about the joists a lot and i too would be worried that if i removed them then the roof might spread. there are two massive wall plates either side of the room , one on the gable end and the other on the opposite wall, these would probably help but perhaps wouldnt be enough to prevent sagging in the middle of the rafters around the purlins. It wouldnt surprise me if the new joists have become supports over the years. perhaps if i did go down this road i could fit in some new/reclaimed collars that would be more fitting and within keeping. the room from gable end to internal wall is 10 feet so its not a huge span so i might get away with one or two? providing i can find a good fixing for them.

Rob - it's important to understand the difference between a collar and a truss. From your photos it looks as though your roof structure is tied together with trusses - that is joists that run across at roof plate level, tying the plates together to prevent the weight of the roof on the rafters from causing spreading.

Collars tend to be placed higher up the rafters, often joined to purlins. As I understand it, collars are designed to prevent the rafters from bowing inwards, due to the weight of the roof. If you were to remove the trusses (ie: ceiling joists) and insert collars, you could radically alter the stresses on the whole structure.

Unless the roof is tied together at plate level, even if collars are in position there can still be a spreading effect on the plates - unless the roof was originally designed with enough rafter strength for collars to act as trusses. As you'll appreciate, if there are collars partway up the rafters, but no trusses at plate level, there can be considerable stress placed on the rafters at the point where the collars (or their mating purlins) join the rafters. The spreading effect below collar level can cause rafters to break at the point where the collars meet the rafters - unless, as I said above, the roof has been designed for this configuration.

That's all a very rambling, amateur explanation, and if it makes no sense, I'll post a quick sketch. So do let me know whether the above is completely unintelligible.

Middi - do you agree with the above? I'd value your much more expert view.

 

[robgil]

i always thought a truss was a construction consisting of say a king post and tie beam? i cant have trusses because my walls are 2 feet high, my tgie beams are in the walls.
edit - a truss will be under tension and collar under compression.
sorry for the short reply , baby feeding time.

 

[robgil]

right then shes fed and playing with Mr B her musical teddy bee with the butter fly umbilical chord.

Penners a diagram would be very helpful, i need to get my head around this. from what i understand as you say a collar will stop sagging and a tie beam will stop spreading, a purlin will also help stop sagging and will eliminate the need for a collar on each rafter.

because my top floor is in the eaves my tie beams are in the internal walls, one at each gable end and one in the middle which has been cut through to make a door way.

i think the joists have been put in simply to replace what may have been there originally which would have been pretty unsubstantial joists to support a lath and plaster ceiling with a more sturdy collar in the middle, or perhaps two.

if these collars have been removed which , if they ever existed in the first place have been , then to remove the joists i haver first to reinstate the collars.

surely there would only be stress on the rafters at the point below the purlin of the walls starte3d to spread under the weight of the roof? if this is the case i think the tie beams in the timber frame should hold things together.

this is the door splitting a tie beam,

you can see the other tie beam in this room t5o the right in the photo,

 

[middi]

Right, Penners description regarding roof members constuction is totally correct.What Penners will not be able to "categorically" confirm is the "safety limit" of a roof that is constructed as per your example. Rooves constructed in that manner were very common ie: rafters, collars, tie beams, purlins, collar clasped purlins, no collars, no purlins etc. These roof structure components were always present and arranged in a manner which is quite obvious ie: stopping the roof from collapsing either inwards, outwards ( at the rafter plates) lateral movement or being blown away by the wind.
These construction designs were invariably copied from previous construction standards.
Now, this is where the "crunch" comes into play. There was NOT( and still isn't) a construction stress calculation standard for oak timber buildings, the builders of the day used their own interpretation ( and the deep/or hallow pockets of the owners) as to what was a "safe" building . Some builders put the emphasis of strength on (ie); just the sturdy wall plates and purlins with not so substantially built rafters and perhaps with just one or no collars. In other cases the rafters may be also quite small but each pair may have had collars which it may or may not have had a purlin incorporated. This latest described arrangement would have been the cheapest to build. The best, and strongest roof constructions were the ones which were intended to bee seen and diplayed to public view, these had incorporated ( dipending on building customs and era) over engineered "beams". Your cottage appears to be well constructed, generally. The roof construction I believe was just intented to serve the purpose of holding the thatch and not collapse , by well educated guess work construction.
I hope this helps
Midi





































.

 

[robgil]

Thanks Middi , so without actually being here looking at it how would you ascertain whether or not it would be a good idea to remove the new ceiling joists and replace with a couple of collars?

 

[chapel]

I went through pretty much the same dilema last year in our early semi detached 17c cottage.

There was a 'more recent' ceiling about a foot above the purlins. The joists for it had been spiked/nailed into the rafters - the wood was quite clearly newer than the rafters. A few of these were wobbly and clearly they werent doing much to help anything (least of all the ceiling which hung from them which was moving/failing).

We took the ceiling and joists down but only after satisfying ourselves that everything would be just fine. There were a few things which swung it:

1) Joists were wobbly and not structurally helping
2) Purlins are great lumps
3) Rafters were good strong lumps (some i think had been replaced a hundered or so years ago - and all pegged)
4) One one side, the rear of the house, the roof slopes down and meets another roof at right angles to it
5) No evidence of original collars
6) Good tie beams


It seems from your pictures Rob that there were/are no original collars - following this logic says then you can take the joists and ceiling out as it was never meant to be there. However......i would say both your rafters and purlins look thin (compared to ours) and the distance between purlin and apex seems seems quite a lot. A lot can happen over the years and you need to be satisfied that the roof will be ok without the joists.

Probably worth you checking around up there, see if there is eveidence of original collars. Check to if the joists have been nailed into the rafters.

Having said this - i'd be tempted to get someone who knows to come take a look....

 

[chapel]

Sorry, just re read by message - it offers no help! Just the ramblings of a mad man

 

[Penners]

Thanks Middi , so without actually being here looking at it how would you ascertain whether or not it would be a good idea to remove the new ceiling joists and replace with a couple of collars?

Rob

First in answer to your question to Middi - you can't tell. And, as I said above, I certainly wouldn't risk it. I emphasise again - trusses (i.e. your ceiling joists - probably) are in tension and collars are in compression, as you correctly stated above. It therefore follows that you can't just assume that a collar will take over the work of a removed truss*.

Timber-frames are frequently hacked about without regard to the stability of the structure, and yet they remain intact. Unfortunately you can never guarantee this. So the only safe approach is - don't hack.

As you requested, I've knocked up the following very rough sketch. It shows in the simplest possible terms the work that each member does. As you observed, a purlin is often used to reduce the number of collars required, so where purlins are in place it's likely that not every pair of rafters will have a collar.

Likewise, a strong roof plate member may reduce the structural requirement for a truss or tie-beam at every pair of rafters. But the fewer trusses that exist, the greater is the risk of spread and possible breakage of the plate beam.

I have also shown you what can happen if trusses are removed - the feet of the rafters can start to spread, the collars start to behave in tension and the rafters can break at collar level.


* Yes you're right, the word "truss" can mean a number of things, among these different meanings it can mean a tie beam - a member in tension that holds others together.

 

[middi]

....(to continue from my previous post). When the roof construction consists of rafters, collars and scantly built purlin, and as Penners quite right indicated, the collars will stop the " inward" presuure but the whole roof pressure is still on the wall plate. This pressure will have the tendency of spreding the roof plate outwards resulting in a bow in the wall plate and a concave or sagged centre section of the roof. This bending of rafter occurs between the collar and roof plate ( the weakest point) and this pressure is concentrated where the collar intersects the rafter. This effect can be noticed on many timber buildings still standing. This spreading is minimised if substantial rafters and wall plate are used and compleyely eliminated if the wall plate is "collared" at rafter feet level. As previous mentioned the size of timbers play a big part in the degree of movement and as to how they are arranged in the construction. This spreading is also well contained if a substantial "bays" rafter purlin is used and supported at each end to transfere most of the roof weight vertically and not "sag" inwards and downwards where, once again, the resulting spreading pressure would be transfered back to the rafter.
Middi

 

[robgil]

cheers Penners, this is a diagram of whats going on at the moment with the roof. I cant see any signs of previous collars up there but when the ceiling joists were put in they wer4e probably only meant to support the ceiling , perhaps now after some years they are under compression as the roof naturally sags inwards? i think if i were to do away with the ceiling it would be wise to replace at least some of them with something nicer to look at than 2x4 soft wood. they seem to be fixed to the rafters just below the purlin so i would assume they are under compression , if the walls were spreading would see the rafters bowing inwards slightly?

ive got a very steep pitched roof so hopefully that might help with the inwards compression?

 

[middi]

Thanks Middi , so without actually being here looking at it how would you ascertain whether or not it would be a good idea to remove the new ceiling joists and replace with a couple of collars?

Since you ask. this is exactly what I would do. ( CO consent permitting)

1) remove all ceiling and collars
2) insert two 8"x8" ( or 10"x10") purlins equidistant under the rafters and supported at each end to bear the roof load ( vertically)
3) Make solid oak "A" plates ( say: 10" x 10") and screw these to the apex of the roof onto the rafters
4) make "stoppers" ( triangular plates from 6x4 seasoned oak) place these above the newly fitted purlin countersink with a 25mm auger bit, drill and fix to rafters with 12mm coach bolt.Fit fake peg into the coutersunk section. This will stop rafter from a downward slide, the new purling will stop inwards pressure and downwards pressure .
result :- vaulted ceiling.
Middi

 

[robgil]

....(to continue from my previous post). When the roof construction consists of rafters, collars and scantly built purlin, and as Penners quite right indicated, the collars will stop the " inward" presuure but the whole roof pressure is still on the wall plate. This pressure will have the tendency of spreding the roof plate outwards resulting in a bow in the wall plate and a concave or sagged centre section of the roof. This bending of rafter occurs between the collar and roof plate ( the weakest point) and this pressure is concentrated where the collar intersects the rafter. This effect can be noticed on many timber buildings still standing. This spreading is minimised if substantial rafters and wall plate are used and compleyely eliminated if the wall plate is "collared" at rafter feet level. As previous mentioned the size of timbers play a big part in the degree of movement and as to how they are arranged in the construction. This spreading is also well contained if a substantial "bays" rafter purlin is used and supported at each end to transfere most of the roof weight vertically and not "sag" inwards and downwards where, once again, the resulting spreading pressure would be transfered back to the rafter.
Middi

cheers Middi, i can see a tiny little bit of sag in the rafters but nothing that would cause me to panic, i cannot see any spread in any of the walls. So if i did remove these new joists fitting a few collars between the purlins would stabilize things do you think?
it looks like ,in the photo of the other side of the house there are two or three heafty collars and in-between there are some smaller timbers to support the lath for the ceilings.

 

[robgil]

Thanks Middi , so without actually being here looking at it how would you ascertain whether or not it would be a good idea to remove the new ceiling joists and replace with a couple of collars?

Since you ask. this is exactly what I would do. ( CO consent permitting)

1) remove all ceiling and collars
2) insert two 8"x8" ( or 10"x10") purlins equidistant under the rafters and supported at each end to bear the roof load ( vertically)
3) Make solid oak "A" plates ( say: 10" x 10") and screw these to the apex of the roof onto the rafters
4) make "stoppers" ( triangular plates from 6x4 seasoned oak) place these above the newly fitted purlin countersink with a 25mm auger bit, drill and fix to rafters with 12mm coach bolt.Fit fake peg into the coutersunk section. This will stop rafter from a downward slide, the new purling will stop inwards pressure and downwards pressure .
result :- vaulted ceiling.
Middi

Thanks Middi , that mostly makes sense to me , what would i fix the new purlins to and i am not sure what a stopper should look like?

 

[middi]

.....my last post should have not read coach bolts but " stainless steel coach screws"
--------------------------------------

to clarify:-

King post= single roof post ( one at each end of bay) supporting a single purlin located centrally under the roof rafter collars. not used much after the 15th C

Queen posts:- Pair of posts ( two at each end of bay) which in turn support a cross truss that supports either a single purlin ( centrally under collars) or two purlins ( one each side of the roof)

Middi

 

[middi]

The best way to fix big purlins to end walls in timber buildings is to rest them on existing or newly inserted supports a required, using straps, pegs or SS bolts as required or a combination of these.

If it is of any interest, Flyfisher is hopefully coming to see me in a couple of weeks, we dicussed taking photos of the cottage I have just finished restoring for my daughter and hopefully he will be uploading them on the site for you (and everybodyelse) to see .

The roof construction and size was very similar to yours and the repairs and vaulted ceiling is very clear to see.

Alternatively take a trip this way and I will show you around

Middi

 

[middi]

......sorry Robgil( forgot). blocks of wood ( carved if you fancy it !!), screwed to the rafters above the purlin and just resting on it.

Middi

 

[robgil]

Thanks Middi , the photos would be good but i trip to Essex could be organised, its only down the road (depending on which part of Essex you are in).
Would all this mean removing the purlins that are there now? I would rather not take anything away if i could.


edit - this is what you mean isnt it?

 

[middi]

Thanks Middi , the photos would be good but i trip to Essex could be organised, its only down the road (depending on which part of Essex you are in).
Would all this mean removing the purlins that are there now? I would rather not take anything away if i could.


edit - this is what you mean isnt it?

Broadly speaking, yes.

As to keeping your existing purlins, it may be possible to leave those in position and attach new ones onto them. Actually it would probably make the fitting and bolting of new purlins easier.

It is something that has to be decided by carefully looking at the problem. Remember:- " there is always a solution to any problem"

Middi

ps I live in Stebbing (near geat dunmow AND BRAINTREE)

 

[robgil]

The only problem i can see with the addition to new purlins is support on the internal wall side, its a wattle and daub wall and i cant see anything substantial enough to fix to or rest upon. there are studs standing ontop of the door frame that i could fix to but putting additional weight onto a door frame may not be a good idea, it is a pretty substantial door frame though.
I suppose i really need to get up there and have a look to see what the original purlins are supported by on either end, i take it they would be supported by something?
would it not be an idea to add the A plates and stoppers , leave the existing purlins then strengthen the purlins with one or two collars?
That is providing the purlins have a good support either end? What strikes me is the structure in place at the moment is working ok and has done for a few years, so removing the joists and inserting a couple of collars between the purlins isnt really altering it too much from its present state. the roof isnt spreading , its trying to sag from what i can see.



Stebbing is just over an hour away from me so i could pop down if you have a free afternoon.