Well, in terms of modern deflection codes under L/180 limits for a roof with no ceilings, most deflection limits would range from 0.50" to 0.75" based on ground snow loads of approximately 40 psf.

It's doubtful larger timbers would exceed these limits, unless their spans are quite long.

Hi Patrick,

We are probably talking more about "creep" here and that is not quite the same thing as a new build "deflections". "Creep or sag" is a long term effect that will more likely take centuries to manifest itself in a negative kind of way. For example suppose you have to replace a rafter in a centuries old church that has a badly deformed roof - would you use a dead straight rafter in full conformance with code requirements or would you try and use the same scantling and deflected form as was currently present in the roof ?

Regards

Ken Hume

Jim –

Some thoughts, never mind the proposed work, has the recent work perhaps already caused more problems than it portends to remedy. I'm guessing the roof was originally wood shingles, which tend to let the snow go with some regularity (mine on the house here, do so two or three times a winter, always with a roar that has first timers thinking the whole house is coming down) keeping the roofs liveload minimized to some degree through the winter. Granted the three-tabs the architecturals are replacing have probably been holding snow the whole winter long for some years, but the laminated construction of the architecturals suggests to me that they also add more dead load per square than any type of roofing the roof system has ever before carried.

Is the new sheathing lighter or heavier than the old ? How thoughtfully was it applied ? I have too often seen historical work poorly re-sheathed, I'm guessing because material takeoffs are done before it is understood how the framing is configured and false assumptions based in ignorance are made. And at times, all the rules of reasonable stagger are thrown aside to reduce the unexpected waste found when purchased lengths do not match real world on-centers, and no thought is spent on how break on break sheathing is concentrating loads to members, members which prior to these “improvements” have to date never been asked so much of.

It may well be the churches frame is carrying more load than it ever has before, and perhaps doing so in ways that it never has before.

And I have to ask, why does the cost of repairing the damaged plaster in the nave fall to the church and its congregation ? It seems to me it belongs to whomever employed those who dropped the boards, or their insurance carrier should they choose to claim it. With potential profit comes potential risk, one of the reasons folks hire professionals is to avoid such risk. Unless there was some negotiated agreement to lower the price, I believe it should not be the congregations problem.

This has been an interesting exploration of engineering in old buildings for me. I need to look into the meanings of creep, sag, and deflection to make sure I understand them in the same way an engineer would think of these terms. I have never defined a tolerance for how much sag should be acceptable in an old frame.

Will;

The reseathing was done with 1 x 6 pine, spruce or eastern hemlock boards and probably all of the joints were broken on the principle rafters rather than being staggared, which I have no problem with in this case. The edges of the boards are tight together months after the work was done so the boards must have been suitably dry when applied.

You make a good point about wood shingles shedding snow and asphalt shingles keeping the snow on the roof. I think the new sheathing did not change the loading on the roof. I have found that roofing nails pull out much easier of new wood than old wood, so I still prefer marginal old boards to new boards.

Being in the trades I hate to see another tradesman take a loss on a job, but I have wondered if the Church committee has discussed holding the roofing contractor liable for the ceiling plaster repairs. I will ask them.

I have been looking into the history of this building and the architect who designed it designed about eight others in the area in the 1830s. Now I am curious to see if any of the other buildings were built the same way.

Thanks;
Jim

I googled some key words and found this paper which gives a good discussion of "creep".

http://www.prugarinc.com/Prugar_my_roof_is_sagging.pdf

Jim

Jim,

If the damage was caused by them they are liable. That's exactly why we pay so much for insurance. I hate it for them, but the ceiling would not be damaged if they hadn't been there...

I learned something important!

I was completely unaware of the engineering term "creep" and so my initial argument that a building that has been around for 100 years or more has proven itself and will keep on surviving, unless some deterioration happens such as insect damage or rot, was wrong.

Creep can continue to happen over time so a sagging rafter can continue to get worse and eventually fail depending on a number of factors in addition to rot or insects, therefore I can't argue that a sagging rafter has equalized or bottomed out and will be fine.

A sagged rafter may be structurally acceptable, but it should be monitored to see if the sag is growing. Of course, there are the asthetics of sagging rafters as another issue, too.

Many thanks to the professional engineer Janet Kane for returning an email I sent her with a lengthy email explaning creep to me.

Many thanks to you too!
Jim

Some follow up questions on creep. It sounds as if it is progressive.

At what rate does it progress? If you went back 50 years, or half the life would you find half the creep? Or does it get worse as it ages, it creeps more as it ages, if you went back 25 years would you half the creep or 3/4's. Is it going to get worse quicker now that it has deflected? Or is it going to creep at a slower rate?

The cause of creep, an undersized member, live and dead loads, gravity, the slope of the roof, anything else? I am curious if heat has anything to do with the cause, roof members are subject to greater heat, does the majority of creep happen during the summer or winter with snow loads?

I presume creep can be seen in other areas as well such as in a post with a long knee wall, bending the post above the tie. You would also see this in the outward bow of the top plate in the same building.

Tim

Hi Tim,

You pose excellent questions and its not surprising that only a few really long term steady state studies have been undertaken in this respect. I know that TRADA in England did a 10 year continuous load test on a truss and then tested it to destruction and found that there was no appreciable loss of strength in the truss.

The members that are most vulnerable to creep are the long skinny ones like rafters, posts or beams where loads applied to same result in measurable deflections. A measure of this effect is determined by dividing deflection by length giving rise to an approximate measure of "strain" (providing the beam is elastic).

Any structure subject to creep is in effect trying to stress relieve itself and as mentioned by me many times on these pages it is important to have secondary load paths available that can take the strain as a structure relaxes upon itself. This is sometimes referred to as "beneficial creep".

In simple terms deflection is dependant only upon load in an elastic structure whereas creep is a measure of deflection that is also time dependant for that same load.

The various variable loads and conditions that you mention above will play a part though it would be very difficult to quantify the effect of each of those individual factors. Unfortunately really old structures (500 years +) that have deformed badly also tend to have had other factors at play in arriving at their current day deflected form in addition to simple steady state loads.

This would form a good dissertation topic for Methusalum.

Regards

Ken Hume

I left out another aspect, a green stick is more noodle like than a dry stick. It seems like a stick of wood goes through various stages, green/flexible, dry stiffer, old and under load/creep. Creep won't spring back, it comes across to me as a memory which is installed.

I think this statement bundles it nicely, Ken.

"In simple terms deflection is dependant only upon load in an elastic structure whereas creep is a measure of deflection that is also time dependant for that same load."

Things just get tired.

Tim