strut and tie modelling

[monchie]

Hello,

I am studying the concept of strut and tie method and modelled an rc wall and try to figure out the correct truss model for the STM.

I have the following questions;

1. How do you normally size the strut? looking at some examples, if the point load is coming from the column, it is quite easy just simply relate it to the size of the column, but how about if you have a "line load", I've just converted the line load to a point load(pls. see attached.

2. When you have a line load, and you intent to convert it to 2 point loads on top of the rc wall, normally, you just take half & half approach(pls. see attachment, in the line load diagram, I enclosed in "red" line load converted in to point loads). But how about the other "halves" outside the "red zone" do I need to include them as well?or just simply leave it according to my model?

3. Based on the result that I've got(Black (Tension), Green(Comp), Red(Reactions)). How do you provide "matching" reinforcement for this?

Any ideas/suggestions are highly appreciated.

 

[Teguci]

1. One way is to have regularly spaced nodes along the loaded edge with equal point loads applied based on that spacing. The way you have it is fine except that the left and right edges should have a vertical strut to carry some of the load. As shown, you are being more conservative. One test is to check if the reactions on your supports are the same as modeled vs with a line load.

2. Do not just take half and half. Loads should be proportioned based on the nodes tributary width (half the distance between adjacent nodes - add left and right nodes).

3. You are working backwards. Pick where you want reinforcing first. Then create a strut and tie model with the ties matching your reinforcement locations. In your case, the right side reinforcement would probably placed horizontal with a hanger/corbel bracket at the dapped end. The STM will need more nodes here to include this flow of tension. In concrete, it is usually cost effective to benefit ease of construction over less material (concrete and rebar).

 

[KootK]

I don't normally advocate non-orthogonal reinforcement but, in this instance, I think that it might be a worthwhile simplification.

I'd assign the distributed loads to the nodes as I've shown below. With the uniform loads at the far ends, you just want to set things up so that those loads are accounted for in the nodal and bearing checks at the reaction points. What I've shown will be a little conservative for moment and shear but will prevent you from having to deal with the complex C-C-T nodes at the bearing points that you'll end up with if you add extra members to the ends of the truss.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Teguci]

Some more rules of thumb -
1. struts run parallel with cracks. If you know or expect where a crack will appear, run your struts parallel with that expected crack. In this case, we know how dapped ends crack so we don't put a strut running from the top node down and to the left in the dapped area.
2. Corollary, struts do not run across cracks. Ties will need to cross cracks.
3. Corollary to corollary - struts and ties can't be in line next to each other. Struts shrink and ties stretch and concrete's shear capacity is finite.
4. Sum of external forces and moments will be exactly resisted by the internal forces and moments at every section along an STM. This can help locate nodes in tricky areas.

 

[KootK]

In the model that I posted, the leftmost vertical tie could probably be omitted as I expect that the loads would tend to rack the truss to the right. If the loads tended to rack to the left, then the compression diagonal would flip directions and the leftmost vertical ties would remain and the rightmost vertical ties could then be omitted.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[KootK]

I would also be perfectly happy with either of the two simplifications shown below provided that you had some healthy crack control reinforcing in the bits of concrete outside of the truss proper.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[KootK]

3. Based on the result that I've got(Black (Tension), Green(Comp), Red(Reactions)). How do you provide "matching" reinforcement for this?

Were I to attempt to reinforce your STM, I'd do something like what I've shown below. I do think that there are more advantageous models for you to use however (we've covered them above).

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[monchie]

Thanks Teguci and KootK,

This is what I'm praying for, a no-nonsense site that provide practical and carefully thought out answer to an amateur like me.

Thanks guys, the info that you gave me is far better than a week of useless reading to all the reference books that I could find on the internet.

Probably, it will take me a while to decipher all the informations that you guys have given me but I will have my comments/queries if I have internalise all of this information.

Again, my sincere appreciation(there's still people that really cares..)


Thank you very much.

 

[monchie]

Hello Guys,

Pls. see attachment.

I am checking the stress at the node to eurocode but I do not know what is the size(area) of the node.

Normally, where do you based the size of the nodes?

 

[KootK]

Normally, you size the nodes to be as large as they need to be for the nodal stresses but, otherwise, as small as they can be so that reinforcement positioning doesn't get whacky.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[monchie]

Thank you very much for your reply KootK.

Pls. see STM result attachment with my comments on it.

I have this situation where this rc wall is supported by intersecting wall at the corner(200mm thk rc wall), the reason for the dapped end. The bearing area is only 200x200. Doing the calculation for the stress in strut/nodes against the allowable (Eurocode), it fails.
Is there any other options to correct it(ie, change the truss model, providing steel plates etc...)

Again, any ideas are highly appreciated.

 

[KootK]

You can add reinforcement to a strut but that won't help you if your problem is really just old fashioned bearing stress at the reaction. I'd recommend this:

1) Check bearing stress at the reaction and report back.
2) Post a sketch of the note under consideration so that we can review what you've done there.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[KootK]

Is there anything preventing you from moving your right end bearing point down to an elevation 2027 below the top of the wall?

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[monchie]

Hello KootK,

These intersecting walls are precast walls and they are supported by 200x200x16SHS via steel capping plates.The reason for the dapped end connection.

Pls. see attachment.

 

[Teguci]

Would love to see the dap go just from a visual ascetic. Can you use something like a double-tee to pocketed spandrel connection? If stresses are too high, you can consider cantilevering a tube steel (HSS4x12 or so) welded to an embedded plate to make the connection to an embedded bearing plate in the hidden pocket of the other wall section.