Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations MintJulep on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Reinforce Existing Joist 1

Status
Not open for further replies.

RFreund

Structural
Joined
Aug 14, 2010
Messages
1,885
Location
US
This is an existing 20H7.
My first question was whether or not the weld shown with a "?" in the attachment can be made and how to denote/design it.
However my second question would be if you have any other elegant solutions for reinforcing the existing 20H7 (Top, Bottom and webs all require reinforcement).
Sorry for the crude sketch the reinforcement would actually be on both sides of the web not only on one as shown.


EIT
 
Wouldn't it be easier to weld a flat plate across the bottom of the joist and use a flare bevel groove weld to weld it to the existing bottom chord rounds?

 
Never mind. I was thinking that your plate was there for flexural reinforcement. I see now that you are attaching your web element to the bottom chord.
 
The weld can be made everywhere except where the new chord interferes. And they might be able to sneak a weld rod in form either side there. Make the 3/16 inch plate big enough that you have some excess on each side to weld to.
Professional welders are magicians. Offer them a six pack of beer and they'll weld the snot out of it!
 
Jed - So the weld between the 3/16" plate and bottom chord bar can be made. That plate would actually be on both sides of the web bar strut and would be continuous. It would be used as reinforcement for the bottom cords as well. When you say where the new "chord" interferes do you mean the new "strut"? Maybe I'm not following correctly.


EIT
 
Rfreund:
WOW... If that’s the only weld or detail on that sketch that you have a question about, you better rethink the whole concept. If you need to reinforce that 20H7 for moment out in the middle, or for deflection, will you have enough shear (reaction) cap’y. at or near the bearing ends? Do you also need to reinforce the last diagonal and the bearing seat detail on the existing jst.? How many of these are there? How long, what loads, moments, reactions? What you’ve got is good for about 500"kips, 5.4kip max. reaction and I = about 185"^4. You have to shore each of the 20H7's before you apply the reinforcement and do any welding. Right now you have a mixed up mess, such that you will never be able to follow a clean load path from a new element to an existing element, at any welded load transfer point.

Why not think of this more in terms of needing to add some moment of inertia, maybe some stronger end diags. and bearing cap’y.? You need to add some top and bot. chord material, and you probably need to add some new diags. to tie them together, since the existing diags. are tough to analyze and pin down, you don’t know their welding either. The three will share the total load in proportion to their moments of inertia. So, let’s make two new light trusses, one for each side of the existing 20H7, and each will have .6(needed Mof I) 50 - 60% of the needed additional moment of inertia. Then we’ll also need to make a stab at the shear flow (welding) needed to tie the new and the old together, to act as one structural unit.

The new trusses will be identical, except one left and one right side. I’m just pulling sizes out of the air to draw my word picture for you, you have to actually size things, check the stresses, and design the welding. The bot. chord on the existing jst. is made up of round bars ~.78" in dia., but the diags. get smaller in dia. toward mid span, correct? I think your .95" dia. must be the end diags. You may want your new diags. to vary in a similar way, to pick up the increasing shear. The new trusses are going to be about 19.5 or 20" deep; to fit under the existing top chord angle horiz. leg and tight against the existing bot. round bar, but still allow shoring. Lay down a .25" x 2 or 2.5" bar for a bot. chord, lay down a .5" x 2 or 2.5" bar for the top chord. These t&b chord bars should have mill edges, not sheared or flame cut edges. These chord bars are laying flat on an assembly table. Use square bar stock or round, and bent as the big boys do it, for your diags.; they could just be straight pieces too. Lay these diags. on top of the t&b chord bars, and weld them together. Your diags. on the bot. chord bar must fit above the existing .78" rounds, and must be below the top of the top chord bar. Your top chord bar may have to be heavier out near the ends for shear/bearing considerations, as may your last couple diags. These new trusses get rolled into place, top first, on either side of the 20H7, with your new diags. nearest the 20H7. Your new bot. chords gets stitch welded to the existing .78" rounds, down-hand and from the opposite side of the truss. Your top chord bars get stitch welded over-head to the existing top chord angles horiz legs.

If you are trying to pick up a new roof load in a specific location (over a specific length, near mid span), you could provide a top chord member which will take an axial loading from each end. Then just put two vert. chord members (compression strut members) at your new load area; then just harp a bot. tension chord member under these struts and up to the end bearing assemblies. Maybe this bot. chord tension member has a turnbuckle btwn. the two struts. These little post tensioned trusses get tied to the existing 20H7 for some stability.
 
DH- Awesome, thank you.
Attached is a sketch of how I interpreted your post. But I am still curious about that weld.

Some background - This is for a new RTU as part of an existing building remodel. The RTU is over-stressing the joist. Actually when I consider the drift load the joist is already over-stressed. I am reinforcing the joist for the RTU and snow load. This will only be effecting a few joist. I will need to increase bearing capacity and will add plates at the bearing locations.




EIT
www.HowToEngineer.com
 
 http://files.engineering.com/getfile.aspx?folder=d3018102-40c9-4ff1-811f-e67e3bc76616&file=Details_-_H-Series.pdf
RFreud,
I don't understand the sketches. Am I missing something, or is the exisitng bottom chords missing fron the sketches? In the first sketch, what is the 3/16" plate doing. Is it just there to connect the web reinforcing to the original chords or is it continuous and reinforcing the chords? Does the existing joist have double angle chords and crimped angle web members? I really don't understand the second sketch. Does it have round bar reinforcing at the bottom chord only, or are the 1/4" plates continuous?

The Steel Joist Institute has a Technical Digest 12 on joist modifications. Much of this infomation is also in some online articles:
Evaluation and Modification of Open-Web Steel and Joist Girders: Strengthening Open-Web Steel Joists
 
W_SE - Thanks for the references and those are usually very help however in this case my joist geomtry is different as it is an old joist.

I should have taken my time and produced a better sketch. The existing consists of Double Angle top chord 1.75x1.75x0.1875" and (2) 0.78" diameter bottom chords with a 0.985" diameter bar web member (varies in diameter).

The first sketch looking mess now is using a bars to reinforce the top chords, a plate to reinforce the bottom chords and an angle to reinforce the web.

However the second sketch is a cleaner and as suggested by DH. It uses 1/4" plates to reinforce top and bottom chords with a round bar for the web. In this process a 'new' truss is made then 'laid into' the existing and attached.

EIT
 
RFreund:
You got the right idea, with a couple exceptions. I hope you are still saying .25" bars in the same generic way I did to paint my picture. You must actually size these, of course. Your new truss diags. must be low enough at the top to miss the fillet radius on the top chord angles, like you show the one on the left. You do show them high enough to miss the round bot. chords on the 20H7, so good there, you got the tight fit-up to the sides of the round bars. You clamp these together to make the welds in the field. Your new bot. chord bars may have to be thicker than .25" but their bot. edge should be above the bottom of the bot. rounds so you can shore up under the 20H7 prior to installing and welding your new trusses in place. Do as much of the welding as you can down-hand and on the assembly table. The top chord welds might just be over-head fillet stitch welds as a function of the shear flow you need. The bot. chord welds might be stitch welds too, but made from above and down-hand. The left side weld btwn. your .25" bar and the round chord is made from the right side and visa versa. Show the weld symbol on the right side, then point through the trusses to the left weld. And, that symbol is a straight line and an arc, not an L. RE: your note 1/4" Pl. T&B, we don’t know that yet, your new bot. chord is primarily a tension member, while the top chord is in compression, so you may want it to be thicker. They both have to be wide (deep) enough toward mid depth of the truss so you can do the diag. welding without nicking the chord edges. I don’t particularly like those flare-bevel-groove welds so I try to avoid them if I can. Square bars for the diags. may cost a bit more, but they will be easier to fillet weld and inspect. Those groove welds are very difficult to make well and consistently, and to know what throat you’ve got. If you make your top chord bar the right thickness and the inside depth of the existing bearing seat (inside of top chord angle to top of bearing pl.) you might be able to just push and roll your new truss into the existing bearing seat depth and weld it in place, no need for any added bearing pl.

RE: your original detail.... I’m not sure you could get your 3/16" bar down btwn. the round diags. and the round bot. chord bars. That space is filled with some nasty groove welds btwn. the two during the original manufacturing process. I’d have to look at some of the bar jsts. to see what that looked like. But, lay out the .78" round bar and the 3/16" flat bar full size, don’t bevel the bottom, and you’ll see they meet at about 10:30 or 11:00 on the round bar; so that weld wouldn’t be bad at all. The thing I didn’t like about your original arrangement was that your new top chord, diags. and bot. chords were not well connected to each other, except through elements of the 20H7, which may already be highly stressed, without the new weld connection stresses; also didn’t like all the difficult field welding. I don’t like real deep flare-bevel-groove welds because you have no control of the root conditions or what the throat size is. If you section a weld like that, as often as not the root looks like a string of chicken sh.+. I doubt you’ll find this in the AWS code, but I’ve put a small bar or solid welding wire down in the groove, then made a light first pass to it and each of the base metal parts, then I can at least be fairly confident of my root condition and throat thickness. This kinda acts as a backer bar.
 
DH - Thanks again.
RE: sizes - yes sizes are just for 'nomenclature' I will size accordingly. The calc part seems to be the 'easy' part here (labor intensive but straight forward enough). I just could not think of an 'elegant' configuration.
RE: shoring consideration - good call.
RE: weld symbol - yes my handwriting is bad and gets worse with the use of "DigiMemo" (google it).
RE: top chord depth - good thinking.
RE: original post - I see what your saying and thanks for the info.

One additional question - when I calculate the effective throat per AISC they use 5/16*R where R is the radius of the rounded member. Say R=0.75". Then the on the bottom left side of the weld symbol the weld size/geometry would be shown as 0.75 (0.23) where 0.23" is the effective throat size and 0.75 is the the distance from the plate to the 90 degree tangent.

Thanks.


EIT
 
I know you are trying to reinforce a joist, and don't know clearance issues you may have, but could you just sister the joist? If you can work out a way to do it, it would be fairly easy. I'm assuming this is not possible if you are looking at reinforcing, but if it is, maybe its worth a look? The joist will take load based on relative stiffness if they are right next to each other, and maybe in this case you add a new joist on each side?
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top