I would imaging there is a ridge beam that supports the peak . How long of a span is it?

"Collar Ties" are not required with cathedral ceilings.

What is required is that once your ceiling joists/cross laterals that hold the bearing walls together are raised above the top plate of the bearing walls....

The ridge must be designed as a bearing girder as wianno previously stated...

AND, the rafter spans need to be reduced and the rafter depths increased as a ratio of how high above the top bearing wall plate they are raised.

Simply put, once you raise the ceiling joists higher than the top plate, the rafter span must be shorter and/or the depth of the rafter increased.

You need an architect or structural engineer or building code official to figure out what you have and what you need here.

A home inspector and most contractors won't have any clue.

Thank you for the clear explanation. I wonder if I should ask the town's building inspector to take a look. It's a rather backwoods/backwards town (central MA), so I don't know what I'll encounter there.

Manhattan42, why do you think the contractor wouldn't have a clue? When I described the issue (the walls seem to be spreading or to have spread), he seemed to know just what I was talking about. He said he would come have a look to estimate what it would cost to stabilize the shifting or to restore things to what they once were "or should have been." I've worked with this contractor before, and I think very highly of him.

My father, who has 2 engineering degrees and is a retired home inspector, spotted the issue right away. But I can't really negotiate with the sellers on the basis of my dear old Dad's observations.

If this is a true cathedral ceiling without any ties there must be a beam at the ridge large enough to support the load of half the roof. The rafter spans are not affected and therefore the rafter sizes are not affected.

Measuring the distance between the walls at different locations, stretching a string from the top corners of the walls, and checking the plumb along the walls can give you some clues but it won't prove the absence of a ridge beam only that the ridge has settled. Sighting along the ridge outside might confirm what you find inside.

I assume you have already looked at all building permits issued for this address at the building department.

Adding a beam would probably be expensive. Adding ties would be cheaper. I saw it done with aircraft cable once but it certainly wasn't code compliant.

This kind of roof, if steep enough, usually settles in the middle until the roof sheathing starts to resist further movement. That's not a solution, it just explains why the roof might have been that way for a long time without failing.

It seems to me that if it turns out that the ceiling was modified without a permit the seller will have to make the repair or reduce the selling price so you can make it.

"And--not surprisingly--the walls seem to be bowing slightly outward in this area."

You need to confirm this observation.

mightyanvil offers some great suggestions on how to do this.

Is the area at the cathedral ceilings finished space? If so and there has been some deflection, I would expect to see some signs in the finish surfaces. i.e cracked plaster, wall board joint separations, open joints at mouldings, etc. Look where you'd expect to see thiese things and check for patches/joint compound repairs or other signs of past distress.

you don't say how old the house is or whether the ceiling configuration was as-built or a modfication.

If it's been around for a while in this configuration and there aren't any signs of cracking or movement, then I can't get excited about it.

My 0.02
So the question is--hw worried should we be about either the deal, or the roof, collapsing?

lynnski,

If the walls are indeed bowed due to an inadequate ridge beam, the simplest and least expensive solution will be to use several runs of either threaded rod and turnbuckles or cable and turnbuckles to pull the top plates together again.

The rods/cables and turnbuckles will be a permanent design feature in your room, extending across the room at the height of the walls' top plates. A place to hang laundry, perhaps? Or at this time of year, holiday decorations?

A good builder can design and fashion the repair, but it's likely that your building inspector may require the solution to be drawn up by an engineer.

Supporting the existing beam with a second placed right under the existong one is a middle option, and replacing the existing beam with a beefier one would be the most expensive.

Mongo

mightanvil erroneously stated:

[quote]"If this is a true cathedral ceiling without any ties there must be a beam at the ridge large enough to support the load of half the roof. The rafter spans are not affected and therefore the rafter sizes are not affected. "[/quote]

While this may be true in your Massachusetts, it is untrue wherever the International Residential Code is used.

The IRC requires a derating of the spans of rafters whenever the ceiling joists are raised above the top plate of the bearing walls.

The higher the crosslaterals or ceiling joists above the top plate of the bearing walls and the closer to the ridge beam they are located, the shorter the span of the rafters can be....

The ridge board must also be designed as an actual ridge girder anytime the ceiling joists/crosslaterals are raised above the top plates of the bearing walls under the IRC.

What mightyanvil states might be true in Mssachusetts, but what he states is a code violation wherever the IRC is used...

The 2003 IRC requirements regarding rafter ties and ridge beams is not easy to understand. It is wise to test the wording against accepted engineering practice in order to avoid narrow contradictory interpretations like applying a Rafter Span Adjustment Factor to a "simple span" rafter between structural supports.

If anyone doesn't understand what I am talking about, here is the clearest explanation I can think of at the moment:

For a simple wood-framed steep-sloped roof structure with opposed rafters meeting at a non-load-bearing ridge board and the other ends supported by and connected to the top plates of the exterior walls, the top plates must be restrained from moving outward by one of three methods:

1) The top plates can be connected together DIRECTLY by structurally continuous joists or ties located at the level of the top plates and connected directly to the plates & rafters

2) The top plates can be connected together INDIRECTLY by an "assembly" (R802.3.1, line 3) consisting of structurally continuous joists or ties attached to rafters above the level of the top plates.
Because the lateral restraint force takes an INDIRECT rather than a DIRECT path between the top plates, additional bending moment is induced into the rafters and they must therefore be increased in size using the Rafter Span Adjustment Factors shown below the standard rafter tables.
(note: The 2003 IRC says these rafter ties must be located at least every 48" horizontally but because that would make it necessary to either increase the size of rafters only at every 48" or increase all of them even though most would not be tied, the IRC has since changed the tie spacing to be at every rafter)

3) Where no DIRECT or INDIRECT connection is provided "at the top plate" (line 16) by methods 1 or 2, the ridge must be supported by a "girder designed in accordance with accepted engineering practice" (line 18).
Since the rafters would then be supported as "simple beams" they should be attached to the ridge girder with joist hangers.
Since there is no additional moment induced in the rafters by any intermediate connectors or ties, the rafters do not need to be increased in size by an adjustment factor and can be sized from the standard rafter tables or by the person allowed by the state to provide "acceptable engineering practice" for the design of the ridge girder, itÂs connections and supports.

If anyone finds fault with this explanation, rather than argue without resolution, post it at the ICC Bulletin Board and see what the members there think of it.

Here is a link that might be useful: ICC Bulletin Board

Thank you all for this great information. I hope to have more clarity after the inspections this week.

mightyanvil--are you really in Mass? This house is about 2 miles south of the Quabbin Reservoir. Wanna stop by sometime?

Sure, how can I find you?

mightyanvil--are you serious? The house is on Beaver Lake in the town of Ware. You can email me at: consume @ yanis-maher dot com. Inspections are scheduled for Wednesday afternoon.

These origninal messages were some time ago but they apply to an addition that I am having an architect design. It will have a 28 foot ridge beam and i am asking that there be no exposed beams other then the ridge beam or support in the middle of the room that would interfere with the view of the lake behind the house. His initial reaction was that it couldnt be done. But with some coaxing he has reconsidered. The width of the roof will be 25 feet. His intial design had two addtion parallel beams but the support on the back end required a post in the middle of our kitchen. Another solution had roof trusses perpendicular to the ridge beam but i again was insistent that there be no cross pieces exposed across the ceiling i wanted it clear. he now says he can do with the one beam (it is massive 28" deep but reading your postings i am worried that i am pushing for something that is going to cause me to have problems later. He is very meticilus about everything he does so i explained that i didnt want him to please me at the sacrifice of a proper solution. i understand that scissor trusses are an option but they would soften the angle of the roof and the height of the windows. Am i pushing for something i am going to regret. Will i end up with bowed side walls?