Mo,
Just a carpenters perspective,
Take a stiff birthday or Christmas card and set it on a table as a "tent" or gable roof. Push down on the ridge, the rafter feet slide out, there's the thrust.

Using a needle and thread make a tie across the bottom and repeat the test. You have made an immutable triangle and proven the tie is in tension since thread doesn't work in compression.

Now move the thread tie halfway up the card and repeat. At the same load the rafters bend and begin to thrust again.

Move the thread up to just under the ridge and load again, I'll bet you have thrust and tear out of the card at the connection.

For the barn in the picture,
If there is just the one tie then it is in tension. In that case the leverage against the tie would make it tough to connect and it would put large bending force on the rafter below the tie. As the tie approaches the plate the forces drop off. I think in the red book it states that a tie should try to be within a foot of the plate. In the IRC rafter span tables it gives a span reduction for conventional rafters as the tie is raised to account for the rafter bending forces.

If you already have a tension tie and then wish to brace the rafters then this raised tie would be in compression. A kingrod with struts might be a better option.

If you do not make a strong triangle of rafters and tie then the ridge should be designed as a beam supporting that section of roof. Yes, the studs are essentially propping the ridge at the gable.

If the tie is below the plate then the thrust from the rafters puts bending forces into the posts, the post is then not just a column in compression it is a column with a combined bending and axial load, this reduces its normal axial load carrying capacity.

Pinegrove's description of the 2 plates with a short section of "tie" between them is neat, never seen that. I guess the pair of beams are helping each other to resist the thrust from building corner to corner. It must be quite flexible to have dropped the new tie, but has obviously held up. I'd sure want a cable or something from rafter foot to foot or turn the pair of plates into a horizontal trussed girder. I guess neither is historic.

Don P. i think you got it in the last paragraph. This is a historic building, am I right Heath? If so here is historic proof of this being done. The done part is the Horizontal Trussed Girder. It's like using a 8"deep by 30" wide plate to resist thrust. This is my suspicion. I like the card test.

As for the studs in the gable ends, if you removed them the wall would not move out the rafter may sag a bit(compression) but the wall will stay put, it has a tie. The middle bent has no tie to speak of other than the theoretical HTG. This building is not all that wide. Heath how wide and long is this building and number of bents? I am wondering if you noticed any bow in the wall at the eve line before you rebuilt it? Did you completely disassemble it, and move it to a new location? Is there a bow now? Tim

Your collar probably fell out because the walls are spreading. I think you need to tie the walls together at the plate level. I would measure from outside of plate to outside of plate and see how much it has moved.

How wide is the barn?

If you need to bring the walls back together, you will probably need to lift the ridge while you pull the walls together. All pretty hairy stuff, unless you know what you're doing. Once the walls were back plumbish, I would install the traditional repair for spreading walls -- tie rods. All approved by an engineer, of course.

Good luck, and keep us posted.

ps I didn't see that several others have already said similar stuff before I posted.

Don's example is a good one. Doing something to resist wall spread is something to think about. Perhaps this was originally designed such that the weight of the roof was considereed low enough relative the wall stability that they didn't tie the walls? Of course everyone will feel better having a tie beam.
One thing that I can't tell from the pic is how the collar tie is attached to the rafters? Was it just nailed up? Is there mortise and tenon and pegs? Calculations could be done using the dimensions, and estimated loads to estimate how much force is being applied to the top plate from the rafters. This would help you size a tie beam and how many and what size pegs to use. Depending on what your intensions are. You could also use a steel rod as a tie if you don't want to go to the truble of a tie beam and don't care about historical accuracy.

the tie was mortise in and pegged


closer look

it was my fault when i drilled for the peg i only had a 1/2 in of meat on the end of tenon.

we had about 4 feet on the roof and 50 mph wind


so i think just setalling of the new hewn side/ we jack and leaved the build put grant back in for a foot so it dropped a little

thank you all for the help

heath enright

I don't see anything tying the plates together in the photo. I would assume your walls spread and failed the connections of tie to primary rafter. It looks like you need to install some ties closer to the plate or external flying butresses.

good day

well the corn barn has two top plats one for the outer wall and one for an inter

their is a tie beam on each side connecting the double plats

i thought a collar tie on the rafter would work

i am just going to put a comalong on the first purlin bring it back togather.

thanks for the help
heath enright
pine grove hewing co

Heath, why not put the come-along in the plates plane? The first purlin will take more force to bring it back. Tim