RAFTER TIE angled cables 4

[Materofact]

not sure why my brain is being thrown for a loop about this. this is why i like trusses by others...

Please see attached picture. basically, an architect wants this look (ya...architects)...... 2 cables for each rafter criss crossing. However, 165 lb factored snow load and 30# dead load per foot on a 24' wide room. Architect wants 8 foot between the rafters! (perpendicular framing between). 8/12 pitch, upper attachmnet of cable is about 4 foot horizontal distance from the ridge.lower attachment is about 1 foot above the plate hieght.

simple calcs (and strucalc type) analysis shows 16k for a typical rafter tie tension at both ends for a connection about 1 foot above the plate, *a little more when you resolve the angled connections.....but with TWO cables the connection should be half of a single member? 10x18 rafters is what i come up with, and about 9kk bolt reqauired connections into the wood. rafters need be designed for the cable loads in addition to the snow and dead, ofcourse. seems huge..?? AND...What about wind thrust INWARD? does roof diaphgragm take care of that once you get it up into the rafter?? how bout stiffening up the ridge joint with a bolted steel gusset?

 

[azcats]

That picture makes me think of KootK. Did he post it here before? I have nothing of value to add here but look forward to this discussion.

 

[r13]

Is the pitch 8H:12V? What is spacing of the rafters?

 

[Materofact]

RETIRED13: the spacing (which is hidden in original text post) is 8 ft centers for the rafters (all other framing ('purlins"?) is perpendicular. 8/12 pitch to me is 8 vertical, 12 horizontal.

 

[KootK]

That picture makes me think of KootK. Did he post it here before?

This one's not mine but I've definitely posted similar in the past. This kind of thing is my wife's wheelhouse so I've got lots of nifty example photos.

simple calcs (and strucalc type) analysis shows 16k for a typical rafter tie tension at both ends for a connection about 1 foot above the plate, *a little more when you resolve the angled connections.....but with TWO cables the connection should be half of a single member?

1) Running through it quick I got about 17 kip unfactored which confirms your number.

AND...What about wind thrust INWARD? does roof diaphgragm take care of that once you get it up into the rafter?? how bout stiffening up the ridge joint with a bolted steel gusset?

2) Yeah, I'd assume that the peak connection shown in the photo you posted is a moment connection for the uplift condition if one exists.

10x18 rafters is what i come up with, and about 9kk bolt reqauired connections into the wood.

3) Watch out that, if you use similar detail to the example photo that you posted, the tension rods will impose a significant moment on the bolt group as the detailing produces a fair bit of eccentricity.

 

[r13]

I think this tying scenario is more effective for roof with pitch higher than 8V:12H, and more prettier/pleasant in looks as shown in the picture. Can you consider double truss and set the rods in between, to create architectural effect and better structural support?
Just speaking loud.

 

[BAretired]

I seem to be ending up with a slightly different tie force when only one tie is used, but close enough for our purposes:
EDIT: it appears to be more than slightly different. It's off by a factor of about 2.

Assuming w = 1560plf, W = 37440#, R = 18720#, Mc = 56160'#, T = 15,178# say 16 kips
EDIT: Oops, I think I made a mistake in the moment, Mc.
Mc = 37440*24/8 = 112,320'#, T = Mc/3.7 = 30,357# say 30 kips

So if a double tie is used, the tension would drop down to 15 kips if the rafters have equal stiffness. If rafters are not equally stiff, the tensions would not be equal.

BA

 

[Materofact]

nice posts kootk and baretired. thank you..... it is still anti- anti-intuitive to me that if you have ONE tie horizontal in the same place (like a normal rafter tie) you would need a connection of approx 16k at each of the TWO attachhments at 1 foot above the plate line like where the bottom cable attachments are placed in this example, but when using TWO total cables(NOT 4 total cables which i think ba thought i was saying?) you STILL need 18k bolt connections at ALL 4 separate connections instead of just 2!!! but hey, THATS WHY do not forego freebody diagrams!!!!!at least thats what i am seeing there from kootk and BAretired? BAretired, im not quite following your FBD. using fixed ends moment calc or simple beam i get twice your Mc.are you just doing one rafter side? so you are also coming up with 16k at each of the 4 connections?

(btw, BAretired a flitch plate is being considered for the rafters which would do what you recommend and also remove all eccentricity worries from the bolt layout kootk brought up).

additonally, with high winds on the WALL, there is "little" to resist the wall from bucking inward since the cables work tension only, however, nobody seems to worry about that load path scenario with ridge beam construction; so i think its overthinking. the plates would need to be designed to span 8 foot between the rafters for that load since all other roof framing will be perpendicular to rafters(or some other way)

KOOTK.... interesting FBD setup.... it seems you are just analyzing one side/rafter load which makes sense, so am i right to say you calc'd that each each of the cable attachments,4 total bolt groups (2 cables per rafter pair just like the pic i posted) requires 18k of bolt connection. NOT 9k like my "intuition" told me because we had 2 cable ties instead of 1.

 

[BAretired]

nice posts kootk and baretired. thank you..... it is still anti- anti-intuitive to me that if you have ONE tie horizontal in the same place (like a normal rafter tie) you would need a connection of approx 16k at each of the TWO attachhments at 1 foot above the plate line like where the bottom cable attachments are placed in this example, but when using TWO total cables(NOT 4 total cables which i think ba thought i was saying?) you STILL need 18k bolt connections at ALL 4 separate connections instead of just 2!!! but hey, THATS WHY do not forego freebody diagrams!!!!!at least thats what i am seeing there from kootk and BAretired? BAretired, im not quite following your FBD. using fixed ends moment calc or simple beam i get twice your Mc.are you just doing one rafter side? so you are also coming up with 16k at each of the 4 connections?

You are correct about the moment, Mc. It has been revised in my previous post to 112,320'#. And the tension has been revised to about 30 kips for a single tie. For a double tie, it would be 15 kips.

If you used a normal rafter tie, 1'-0" above the supports or 7' below point C, the tension would be Mc/7 or 16,045#. It should not be surprising if you consider that a horizontal tie has a lever arm of roughly twice that of the sloping tie.

I didn't draw a FBD. The tension in the cable is the simple span moment divided by the perpendicular distance from the cable to point C.

(btw, BAretired a flitch plate is being considered for the rafters which would do what you recommend and also remove all eccentricity worries from the bolt layout kootk brought up).

Okay, that sounds like a good idea.

additonally, with high winds on the WALL, there is "little" to resist the wall from bucking inward since the cables work tension only, however, nobody seems to worry about that load path scenario with ridge beam construction; so i think its overthinking. the plates would need to be designed to span 8 foot between the rafters for that load since all other roof framing will be perpendicular to rafters(or some other way)

I would be tempted to change the cables to HSS members designed to take whatever compression is required to resist wind pressure. Needless to say, the architect would not like that.



BA

 

[r13]

For purpose of discussion, analyses were performed on a model assuming rigid ridge point. The first graph shown analysis results for DL+SN, and the second, for DL+WL.

 

[KootK]

so am i right to say you calc'd that each each of the cable attachments,4 total bolt groups (2 cables per rafter pair just like the pic i posted) requires 18k of bolt connection.

Yes in my opinion.

...it is still anti- anti-intuitive to me...

Yeah, I feel 'ya. For me, the key is the recognition that you don't actually have more than one rod acting as your truss's tension chord at any location. Perhaps the sketch below will speak to your intuition a little more than FBD's.

 

[BAretired]

retired13,

[li]I'm not sure how you would get a rigid ridge point using timber rafters. I assumed a pin joint for Node C (your Node N4).[/li]
[li]Your load of 0.195 k/ft does not correspond to the reactions shown, i.e. 24*0.195/2 = 2.34k (not 2.812k at N1 and N7)[/li]
[li]N3 and N5 should be 4' each side of N4, not where you show them.[/li]
This makes comparison of our results difficult.

BA

 

[r13]

If the ridge point remained pinned as in the simple truss construction/analysis, without restraint at wall-roof connection, the outward deflection of the legs could be huge. The stiffness of the cables is crucial in minimizing the lateral deflection, the cables need to work together. Suggest to consider making the ridge connection as rigid as possible.

 

[BAretired]

retired13,

I disagree. Making the ridge node rigid has a minor effect on reducing the spread.

BA

 

[r13]

BA,

Why not do a check to see the effect (I did above, in case you've overlooked)? Also, the stiffness of the rafter will play a role too.

 

[BAretired]

I agree that the rafter stiffness will play a significant role in the magnitude of spread, but the rigidity of the ridge node will play a minor role. In any case, moment connections in wood members are difficult to accomplish.

BA

 

[Materofact]

thank you for your continued attention to this post. and how bout the size of them rafters eh??? as long as youve come this far with it!! i was at 10x18 when i thought is was just 9k connections. have to load the rafter withthe cable tension load and the snow! the rafter size requirement seems to indicate the angle of this roof is probably not steep enough for a sensible design? flitch plate will work ofcourse, but big$$$$cash// 2 6x12 with a 1.5 inch flitch plate.?? approaching reeediculous!

 

[r13]

Yes, I guess it is quite difficult to make a rigid wood joint, but remember that it is equally difficult to make a pin too. I am thinking to add a small wedge that is tightly fitted/attached to the rafters through truss metal fittings. I've no way to test it, so just treat it as a noise. BTW, in my analysis, the rafter is 2x12, the tie is 1" dia. rod.

 

[BAretired]

Big difference between a 2x12 and a 10x18.

BA

 

[dhengr]

Materofact:
You are designing/building a scissors truss without any wooden diag. web members, and with the tension members being end-threaded steel rods instead of timber members. There are turnbuckles and end clevises on the tension rods/members. Scissors trusses always have a fair amount of lateral thrust/deflection/movement at their bearings, and you must design for that. The tension rods allow you to adjust this. You say you need 10”x18” rafters, and I wonder if you don’t really mean you need two 5 or 6”x18” timbers, side by side, to make up a rafter. These are sizes which are easier to come by than the larger size, and likely easier to fab. the trusses around. The Arch. is making the design and fabrication very difficult, for little aesthetic affect, by insisting on the 1’ up from the bearing elev. for the bottom connection, and he has misaligned the connections if he does them as shown in the photo (mentioned earlier). I would have the center of the connection pls. and bolt groups match the line of the tension members, thus much simplifying the connection and bolt group design. I would move the bottom connection down and make it and the truss bearing pl. one fabrication. This consists of a .375” or .5” center pl. (some such thick.), it has a horiz. bearing pl. to fix to the wall top pls., and it has a vert. exterior pl. to take the thrust from the rafter pieces. In effect, this is a shoe for the rafter ends, and now all the bolts do is account for a little joint eccentricity and hold things together in the through thickness direction. The top connection could be an angled “tee” shape, to match the tension line, with the top flg. dapped own into the top of the two rafter pieces, and screwed in place, and again, all the bolts do is account for some joint eccentricity and hold things together in the through thickness direction. You could provide a “vee” shaped ridge spacer pl. with a few bolts, for aligning the rafter pieces and a little fixity (not full fixity). A full truss sized flitch pl. is possible, but very expensive and difficult to deal with, and wouldn’t seem to offer much here. However many trusses you have, you have two of the bearing shoe fabs. and two of the top connection fabs, and one ridge spacer, for each truss. But, they are all identical, and this kind of hardware fab. is not uncommon for a good truss manuf’er. The truss guys could also cut end mortices for the purlins in one side of each 6” (or 6+) wide rafter piece, thus simplifying the placement of the purlins. The end fixings would then be a couple 2x2x3/16th angles punched for a few structural screws.

Edit: At the tension rod crossing, provide a nice little saddle shaped piece of hardware to keep the rods from rubbing or vibrating w.r.t. each other; and as a means for hanging lighting, etc. You might add a vert. pipe btwn. this crossing and the ridge. Is that a king post or a queen post?