Beam Reinforcement Detail 3

[RattlinBog]

I'm designing reinforcement for an A36 W12x22 with 25-75% section loss due to corrosion on its top flange and web. The W12 is a strut in a braced frame with about 20 kip axial and is also picking up some floor loads with 1.25 ft trib width. Prior to corrosion, the beam is at about 32% bending & axial utilization; after corrosion, it's at about 122%. The corrosion extends about 5 ft from one end, and the beam is in good condition elsewhere. The 122% utilization may be a bit conservative because there's no section loss at midspan. This W12 would be challenging to replace, so I'm looking to repair.

Anyway, would anyone have suggestions for improvement for my attached detail? I'm welding 1/2" plate to the flanges and 1/4" plate to the web. Might be a bit conservative, but I'd rather keep symmetry. Does anyone see fit-up issues with installing and welding plates MK-9P6 after plates MK-9P5 have been welded? Anything I could do better?

Thank you

 

[dik]

Can you use round bars in the top fillets and a flat plate on the bottom of the bottom flange?

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[driftLimiter]

Also I have seen Angles added in a similar location to the bars that @Dik describes. Your solution requires a lot of preparation for the welds.

Are you concerned with the web as well? those doublers plates look pretty rough to me. If you can get away with just reinforcement of the Top and Bottom flange similar to what @Dik is saying I think thats the simplest. If you have to reinf. the web maybe try to get it on one side only?

 

[SlideRuleEra]

1) ...W12x22 with 25-75% section loss due to corrosion on its top flange and web.
2) The corrosion extends about 5 ft from one end, and the beam is in good condition elsewhere.
3) ...I'd rather keep symmetry.

1) How much is section is lost is not as important how much remains. In this case, sounds like the 7/16" thick W12x22 flange is now less than 1/8" thick in places... not much to weld on.

2) The proposed 6'6" long repair is welded almost exclusively to the 5 feet of corroded, compromised steel. Structurally, the welded repair may (probably will) "look" better than it is since that part of the existing W12 is in such poor condition.

3) Very good... but, because of the axial load, I suggest keep the repaired section symmetric along it's length, too. That is, extend repair for almost the full 21' length.

Because of excessive metal loss, suggest a bolted repair instead of welding. Perhaps like this, with the MC10s butted up against the (assumed) corroded connection to transfer axial load to the column without relying on the W12 connection. Shim between the W12 and MC10 flanges, top and bottom, at bolt locations.

 

[jayrod12]

Honestly for that amount of work, wouldn't it make more sense to just cut out the offending portion and replace it with a full moment splice once you're back to solid steel?

 

[RPGs]

Always tough trade-off going for a splice vs. reinforcement. At least your proposed repair is all fillet welds, but that is a huge amount of field welding, especially if the section loss is that significant--in the act of cleaning the thing might fall apart.

Honestly for that amount of work, wouldn't it make more sense to just cut out the offending portion and replace it with a full moment splice once you're back to solid steel?

If there's access to both sides to do all this welding and you have that much good steel 5 ft away, perhaps a thick end plate (field fillet welded to the cut end with good steel), bolted to a healthy W12x22 transplant w/ corresponding bolted end plate, might be feasible with temporary shoring work. That said, keeping a braced frame stable with temporary shoring can be a whole different ballgame if you're talking about 20 kips of load.

 

[TLHS]

The bolted channel was what I was thinking of. I'd also just consider running new channels or beams on either side and not directly fastening to the old one. Completely replace the load path.

Basically, how confident are you with your understanding of the corrosion of that beam. That's a lot of section loss. If you keep getting failure in that top flange, are you sure you're going to have a load path into the beam in bearing and other similar things? Are you sure you don't have a local issue that could cause you problems in a way that we normally assume isn't an issue because of width to thickness ratios, or our own internal assumptions? Are you sure you haven't compromised something already?

 

[RattlinBog]

Everyone, thanks for the replies. I'm still at work, and our internet security doesn't let the images copy/pasted to forum replies load for some reason, so I'll have to look at those at home.

For reference, I'm attaching a photo of the W12. "WTF HC to TT" = west top flange - heavy corrosion to thin & thru. "HCTT" = heavy corrosion, thin & thru. I see that I had a typo above; the top flange corrosion extends to 6 ft. The web corrosion appears to be localized. This was a tough spot for our inspectors to reach, so they didn't end up taking UT readings. Appears to be a visual inspection. I believe the west top flange is knifed down from 40 years of corrosion. Hard to tell how much steel is remaining...I may consider a splice or an alternative repair that was suggested.

For a little more background, I'm working on a replacing a lot of this floor (beams & plates)because of significant corrosion. This W12 is one of the "good" beams.

 

[RPGs]

Visual inspections are always tough if that's what you're going off of.

Did your inspectors perform any scraping (hand wire brush or wire brush attachment to drill) to get those measurements that you based your section loss on, unless HCTT means perforation through the section?

 

[dik]

I had that problem at the last place I worked... IT wouldn't let me use the wifi... They were not happy that I didn't use their hardware... used my laptop and uploaded my work at the end of the day. For wifi, I used to walk over to Tim Hortons and use theirs.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[dik]

SRE... you need something to bolt to <G>... might be that you cannot use the web to transfer shear, either.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[SlideRuleEra]

...you need something to bolt to...

That's one of the reasons for my suggested near full length splice (say 20 feet reinforcement for the 21' span)... bolting through 14' of "good" web to cantilever 6' of damaged beam. Shims at all bolt locations (probably about 2' on center) to pick up the load.

Of course, if the beam-to-column connection is too weak to carry the load you are right... you can't make a silk purse out of a sow's ear and partial / full beam replacement or added beam(s) have to be considered.

There should not be too much bending moment or shear with a floor tributary width of only 1.25 feet.

 

[RattlinBog]

Thank you for all the replies. I apologize if I don't reply to everyone individually, but please know that I am reading and considering all posts.

A little more context below. I'm trying to walk the line between being too vague and revealing too much info. I'm a structural engineer for an industrial facility, so I'm essentially the client. I want to keep some reasonable level of anonymity out of respect for my company.

I hired a firm to inspect the floor due to suspicions of heavy corrosion. They did a nice job and were able to UT all but a couple beams, the W12 in question being one of them. I spoke to the inspector again yesterday afternoon, and he confirmed that the web corrosion is localized to just a 12" width shown in the photo. I can fix that with a doubler plate. (There is an opening in the steel floor plate for washdown above that beam, so I believe that is how the localized corrosion happened...over 40-some years.) He also confirmed that the west top flange is thin/knifed down. Again, the interface between top flange and steel floor plate plus consistent moisture gave us these results. The rest of the beam is okay.

I like the bolted N.S./F.S. channel reinforcement; however, I do have a concern with the added shims being a new source for pack rust to rear its ugly head. This floor will continue to be wet. I'm a relatively new engineer, but I have seen case after case of pack rust issues in the industry that I work in.

I'm starting to consider the partial replacement/splice option. When construction on this floor begins, about 90% of the beams and 100% of the floor plate will be removed/replaced. On a windless day, I believe this W12 strut will see just self-weight. Normally it sees 15 psf dead + 100 psf live for 1.5 trib width, so about 9 k-ft service moment. The 20k axial is a bit of an estimate based on tension-only forces in the bracing. Perhaps we can cut and splice in a new W12 piece.

Another option I'm considering is to cut out just the top flange for its ~6 ft of damage (once loads are removed) and replace the the top flange with either double-angles, a channel, or a plate. See attachment for a double-angle section. I would fix the web with a doubler plate. Any thoughts on this? I'd like to stick to a partial-length repair or splice; otherwise, I will probably replace the whole strut if I'm looking at full-length reinforcement. (P.S. we have an in-house fabrication shop that does nice work. They can handle fairly intricate steel details. As far as cost, I factor in outside inspection and engineering, construction labor, and materials. If we do our own fabrication, shop labor is null.)

Again, thank you for the commentary. This gets me thinking about different options I hadn't considered.

 

[SlideRuleEra]

1) I'm a structural engineer for an industrial facility, so I'm essentially the client.

2)When construction on this floor begins, about 90% of the beams and 100% of the floor plate will be removed/replaced.

1) Welcome to the club... I worked for an electric utility (generation). I see in another thread you worked 4 years for a consultant. This type work is different, you are not the Client, you are the Owner. The work is not about creating the best, efficient design. It's getting the work done, within the limitations of existing conditions, with the least negative impact on your company...

2) In this case, the answer, IMHO, is a clear no-brainer: replace this beam, too. There are too many unknowns, extent of existing corrosion and how to clean it satisfactorily. Future corrosion. A "brand new" floor with one questionable beam (this one)... what do you do when this beam "fails" in the near future? Possibility that repairs to this one beam will drag on due to unexpected damage... delaying project completion... perhaps impacting plant operations.

Again... replace this beam.

 

[RattlinBog]

SlideRuleEra, thank you for the correction. I'm still getting used to the "Owner" hat. Some days I feel like a mini consulting firm within my own company. (Since you read my previous thread, I'll give an update that I did pass the PE exam and have become licensed as of September.)

I believe you're right. I will digest that for a bit, but that does seem to be the best way to go. Speaking from some lack of experience (relatively speaking compared to the senior members on this forum) do you foresee issues with temporarily disconnecting this strut? I could look at shoring, but perhaps it would be unnecessary if there is no load. There are other struts and bracing above this W12 that I believe would be sufficient for any lateral load during construction.

 

[SlideRuleEra]

1) I did pass the PE exam and have become licensed as of September.

2) I will digest that for a bit, but that does seem to be the best way to go.

3) ...issues with temporarily disconnecting this strut?

1) Congratulations! You will always be glad you passed the exam early in your career.

2) Exactly what you should do, you might change your mind. As Owner, I expect you have a huge advantage over a Consultant... you can easily make as many site visits as needed, and I suggest you do so. Each time you will see things differently. I would often make a dozen visits, or more, on one project. Would tell my Understudy "We have a full-scale, 3-D, model of exiting conditions, let's make use of it."

The 2 of us had technical responsibility for major structural repairs/modification for 7 generating stations with total 3.2 gigawatts capacity (for reference, it took "only" 1.21 gigawatts for Doc Brown to send Marty "Back to the Future").

3) That depends on what causes the 20K axial load (environmental, plant operations, etc.) and how the total structure is designed. It's may be a real possibility, I've done it,sometimes, look into it. See ASCE 37, "Design Loads on Structures During Construction" for some guidance.

 

[RattlinBog]

1) Thank you! It's a good feeling. Now I just need to finish up my M.S. Civil Engineering degree over the next couple years...

2) Yes, I definitely agree on the unlimited site visits advantage. That was one thing I struggled with in consulting. I really enjoyed design and taking part in occasional inspections, but some days I felt like I was designing blindly because there usually wasn't enough budget to make multiple, thorough site visits. Most of my design work dealt with existing structures (like now) and it was challenging to get everything I needed in that first kickoff visit. I digress. Anyway, for this particular floor, access is very limited. Inspection requires a ~40 ft lift and the ground terrain is less than desirable. This floor was ignored for a long time until I got hired a year ago.

3) My understanding is the lateral load for this frame is just wind loading. I have access to the original 1970s design drawings for reference. Using an ASCE 7-16 Ch. 26/27 Mathcad sheet I built for my M.S., I'm getting about 35 psf windward, 16 psf (minimum) leeward, and 11 psf roof pressures. Building dimension normal to wind is 33.5 ft, so I'm getting 590 plf windward, 270 plf leeward, and 180 plf roof wind load based on tributary width for this frame. Attached is an analysis comparison (W12 strut intact vs. removed) I did in a web-based program called Strian. I try not to lean on software too much, but Strian is nice for analysis-only checks, especially for indeterminate structures.

I'm seeing +/- 1 kip axial force difference between the W12 being intact/removed. The braces are 2L2-1/2x2-1/2x1/4, which are good for 51k tension. Worst-case service axial I'm seeing is about 20k. Appears I was wrong about my earlier estimate for the axial force in the W12 strut. These results show about 13.5k. Then again, I was getting about 20k in the W12 by assuming tension-only bracing. It appears I'm ok to temporarily remove this strut. We're not going to be doing this work in a design wind load scenario. However, I'll sleep on it and perhaps decide to sister/shore the strut with a channel temporarily welded to the exterior flange of the two columns for stability.

 

[Teguci]

Have a look at the existing beam for camber. Elements that are stiff against bending (the vertical web plates) will not be able to match the existing camber in the field without some trouble.

 

[WinelandV]

We're not going to be doing this work in a design wind load scenario.

This is where the ASCE 37 comes in - how long do you anticipate the replacement to take? ASCE 37 has wind reduction factors for various construction durations.

Of course, just providing an alternate load path for the full design load has its own advantages.

Please note that is a "v" (as in Violin) not a "y".

 

[RattlinBog]

SlideRuleEra and winelandv--thank you--I will buy ASCE 37-14. I haven't used it before, but it only makes sense that I have it. I can swing $84...even though it's $1.75/page lol. I'm only kidding, though it is sometimes discouraging how much money we have to spend to stay current on so many topics.

Teguci, good point and thanks. I don't always think about camber.