Sagging roof 2

[JStructsteel]

So I looked a small house this am for a roof that is sagging. from the outside you can see the sheathing bowing between the rafters, and some noticeable bow in the rafters. This house is under contract for sale, and the inspector wrote it up. I checked it out, the rafters are 2x4 spanning about 12'-0" from ridge to eave. How do you justify that its been that way for 70 years, the wood is just creeping, and that its fine, when you know the design would never check.

The wood is in very good condition, i suspect the sheating is 3/8" or 1/2" max, and is just creeping over time.

Structurally Its in no worse shape than the 100 other houses built in the neighborhood. Just need to find a way to say in a report.

Suggestions?

 

[KootK]

How do you justify that its been that way for 70 years, the wood is just creeping, and that its fine, when you know the design would never check.

You could attempt to quantify what is almost certainly the reality of why the building remains sound: diaphragm action from the sheathing, particularly if there is celling framing effectively acting as a tension tie between the rafter bearing seats.

 

[IRstuff]

Did the roofing material material itself change from initial? People often go from cedar shake or shingle to cement tile, and that adds something like 3-5 lb/sqft and the roof surface will often show sagging. If they did, then the roofer might have pulled a permit to do so, and you can reference approval of the permit and the post-installation inspection. Is the sheathing full sheets of plywood? Cedar shake is often installed on lathing that matches the roofing course dimensions. Presumably, one can do an analysis of the roof structure to show that supports the wood sheathing and roofing material.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[KootK]

I looked at a very similar setup a few weeks back. A robust, 2x3 ceiling diaphragm exists down at the batt insulation. In this instance, I didn't bother trying to quantify anything. I just said:

1) Here's what you've got and how I think it works.

2) Been there since 1910 and it isn't going any place unless something changes.

3) Revisit this with a pro if you plan to do any renos that would add load or mess with the roof or ceiling sheathing.

 

[phamENG]

Just because it has stood that way for 70 years doesn't mean it will for another 70. Has it ever seen the full design load mandated by current code? Probably not. So be careful in saying that it's ok because it's been ok in the past. Say it's ok because you believe that it is, based on your engineering judgement.

Where I live, houses built after 1954 have not gone through a design event. Even most of the houses built prior to that probably didn't see the design level winds in 1954, as that was measured by a buoy on the water and not over land. And yes, that is accounting for variations in reporting based on elevation and duration of gust. So for me, a house has to be waterfront and at least 66 years old for me to consider the possibility that it was exposed to a design level event.

 

[JStructsteel]

Thanks.
The roof covering is shingles, certainly not original. It is 4x8 sheets, with some repair areas cut in here and there.

Its south west ohio, certainly has seen it share of snow, and its not a huge pitch either, 5 or 6 on 12.

 

[skeletron]

Hand-frame roof rafters? Are there collar ties? Otherwise, what Koot pointed out, the ceiling joists help keep pulling it inward. Make sure you separate the dead load on the actual roof versus on the ceiling. I think it is about 50-50 split, but it can make a difference.

 

[KootK]

Just because it has stood that way for 70 years doesn't mean it will for another 70.

In a probabilistic sense, a past history of successful performance absolutely implies an improved likelihood of future successful performance, even if the design loading event has not been reached. In the Canadian code, we actually have nifty provisions for this that allow us to adjust the safety factors accordingly. That, using the exact same probabilistic logic used to generate load and material safety factors in the first place.

With each passing year, a successful structure is proving its worth to a greater degree.

 

[phamENG]

KootK - certainly, especially as it relates to dead and live loads and other more predictable loads. Reference design values for wood put you down in about the 5th percentile for wood strength. So once it's been built and has proven it can hold the design loads, you've proven you're at least there and likely higher, so a reduction safety factors is reasonable. I'm a little more cautious when it comes to more extreme weather/natural events, though. I'm in a hurricane prone region, and when it comes to saying an existing roof structure or uplift load path is good, I take a pretty close look and typically discount it's past performance. It didn't get ripped off during the first severe thunderstorm, and that's good, but it hasn't seen a category 2 or 3 hurricane, either.

Another thing to think about is load duration and cumulative load effects. OP is in Ohio, with pretty good snow loading. Load duration factor for snow is typically taken as 1.15, which corresponds to 2 months of cumulative time at the design snow load. I imagine a 70 year old structure has eaten into those 2 months considerably, which may undo some of the gains from the probabilistic gains based on age.

And that's neat that you have that built into the code. If anyone is aware of it in US codes, please share.

 

[KootK]

I'm a little more cautious when it comes to more extreme weather/natural events, though.

So is our code treatment of this. Reliability level is based on how long it's been in service and some other stuff.

Brain teaser: if a roof deigned for snow survives another year in which no snow falls, has its probabilistic capacity to resist snow increased over that year? I think so.

 

[StrucDesignPE]

OP says it is "just creeping over time," but what about the possibility of creep-rupture?

 

[r13]

I'll write the report just like your observation - bowing, sagging, no immediate danger (need some calculation here), but expecting re-roofing soon. All facts, no guess works.

 

[Tomfh]

Brain teaser: if a roof deigned for snow survives another year in which no snow falls, has its probabilistic capacity to resist snow increased over that year? I think so.

It has to experience loads above that which it has previously experienced for its predicted capacity to be improved. Its capacity doesn’t improve just sitting there doing nothing.

 

[KootK]

All facts, no guess works.

In a building that's 70 years old, it's unlikely that you'll be able to take the allowable material stresses as any kind of "fact". At least not unless you bring out a lumber grader or send samples off to the lab which is rare for residential work. The line between reasonable estimation and educated guesswork can start to get a bit blurred...

 

[KootK]

It has to experience loads above that which it has previously experienced for its predicted capacity to be improved. Its capacity doesn’t improve just sitting there doing nothing.

I see now that I should have phrased that more precisely.

Brain teaser: if a roof deigned for snow survives another year in which no snow falls, has its probabilistic capacity to resist specified snow loads increased over that year? I think so.

When a additional year passes with no snow, it helps to demonstrate that the original factor of safety used for snow load evaluation was unduly high. In accordance with the relation shown below, that effectively increases the specified load that can be resisted. It's just working the other side of the Demand/Capacity equation.

Maximum Specified Snow Load = Factored Specified Snow Load / Appropriate Safety Factor.

 

[KootK]

OP says it is "just creeping over time," but what about the possibility of creep-rupture?

I believe that the load duration factor takes account of just that phenomenon. phamENG raises an interesting point about the using up of the assumed load duration though. When we say that the appropriate load duration factor for snow is two months, what lifespan are we assuming for the building in that statement? Fifty years? Eventually, most buildings outlive their design lifespan.

My own house is 42 years old. Does that mean that, in eight years, I'll need to reinforce my roof against an eminent creep rupture failure because I've statistically used up more than my allotted two months and the originally assumed load duration factor is no longer valid?

I'd have to do some research in order to even figure how to reinforce for an eminent creep rupture failure. Jack stuff up and reinforce it such that the original members will be exposed to a narrower range of stress in the future??

 

[r13]

In a building that's 70 years old...

Too many possibilities beyond a simple observation can tell. It stand for 70 years is a fact, the bow and sagging are facts, will it be standing on tomorrow after a historical event, through some calculation, it can be a fact by judgement. A warning for corrective action in the future is sufficient, unless the client asks "how long".

 

[Tomfh]

When a additional year passes with no snow, it helps to demonstrate that the original factor of safety used for snow load evaluation was unduly high

Do you mean an additional year of no snow reduces the theoretical design snow load, by reducing statistical likelihood of snow events?

 

[KootK]

Do you mean an additional year of no snow reduces the theoretical design snow load, by reducing statistical likelihood of snow events?

I think so but, again, in the interest of precision I would rephrase that as:

1) The likelihood of the originally assumed, design sow event (specified load) occurring in any one year remains unchanged over time and is purely a function of the climactic record and probabilistic assessment.

2) With each additional year that passes where the actual snow load experienced does not exceed the specified snow load from #1, we have more confidence that the ultimate limit state snow load will exceed the specified snow load by a factor less than was originally assumed when the structure had no available history of performance.

3) #2 Implies that a lower factor of safety on the loads is appropriate.

Statistics isn't really my strong suit but my fundamental understanding of this is essentially this:

4) A successful structural performance history adds information that wasn't known when the structure was brand new.

5) All new information reduces uncertianty.

6) Reduced uncertainty should rationally lower appropriate safety factors since uncertainty is precisely what safety factors are meant to account for.

It's a bit like the classic Monty Hall problem. When a year passes without snow, that's a bit like opening one of the doors without a car behind it. You know more and can therefore guess better.

 

[Tomfh]

1) The likelihood of the originally assumed, design sow event (specified load) occurring in any one year remains unchanged over time and is purely a function of the climactic record and probabilistic assessment.

Our additional year of no snow increases our climate record, which could be used to recalibrate the snow probability model (however slightly), and this in my view is the only conceivable way the probabilities are affected.

It's not the Monty Hall problem. Snow events are not fixed goats and prizes behind doors that you progressively open.