Roof reinforcement

[JRS56]

We're looking at reinforcing an existing flat roof for additional snow load due to construction of a new addition that has roof line that is 5' taller. The existing roof is 100' x 100' with a mid span beam (W section) supporting the 50' long bar joists. Bar joists are 24" deep x 50' long x 48" centers.

Previous building owner faced a similar situation about 15 yrs ago when they added an addition at the opposite end. At the time the previous owner chose to modify the bottom chord of the bar joists by welding a 1/2" x 6" plate to bottom chord. There was no engineering review just a seat of the pants design by one of the key employees. This has worked well thru several major storms over the years.

New owner wishes to use W14 x ? as infill beams between the existing bar joists. Keeping the infill beams shallow allows him to not interrupt utilities that are running thru the existing bar joists. I've designed a grid of 5 pcs W14x43 x 50' long beams with bridging on 12.5' ctrs. to support the top flange of the W14.

The dead load induces about 1/2" sag at the middle and 2.5" at full snow load.

- I'm curious about how to reconcile the sag from dead weight?
- How close should the stiffness of the new beam grid match the existing bar joists to insure best loading practice

Thx in advance

 

[phamENG]

The existing roof system should have some initial deflection from the dead load on it now. A properly designed systems of open web steel joists with full roof dead load could well have a half inch of deflection in it as is. In which case, you might not have enough self weight deflection to get the beams into place easily.

If you want to ensure equal load sharing, you need to ensure equal stiffness. Keep in mind that if you don't jack up the roof during installation, the roof dead load will be carried by the existing framing and the live/snow loads will be split to the systems based on their relative stiffness.

All that said - this is a job for an experienced structural engineer. Replacing or significantly altering the structure of a 10,000ft[sup]2[/sup] roof should be subject to permit review by the AHJ - even it it's on an industrial site.

 

[JAE]

How do you get the new beams up between existing joists when you stated that there were utilities running through the joists? Not sure I follow you.

Are the joists perpendicular to the addition or parallel to the addition?

 

[JRS56]

@JAE existing joists are running parallel to the new addition. To get new beams in place we're thinking of a splice joint at maybe the 16' point. We need to snake the longer length over the existing utilities.

I've submitted the existing bar joists to SJI for any possible historical info. Building dates to mid 70's. Building plans show "24H8" as the bar joist of choice. Joist tags read "H56102-1 R2"

As I mentioned above, about 3 years ago we had the storm of the century with about 7' of snow over 3 days and the older joist modification held up fine with no detectable permanent set.

We plan to submit plans to the local building inspector for review.

 

[dold]

I would be very hesitant to reinforce only the bottom chord of a bar joist (although that's already been done). Bar joists are typically designed to ~99% of capacity. That is almost always controlled by some sort of buckling mechanism of the top chord. Adding a plate is great if you're dealing with a wideflange, for example. The added strength is a result of shifting the neutral axis lower. This is not necessarily true for bar joists. Also note that web members, and even length of welds between web members and chords are optimized and warrant a quick check of capacity.

If you have the building plans with joist designations, you can look up the allowable loads in the historical joist catalog that is published by SJI. Joist tags are (typically) useless, unless you have the joist shop drawings. You can also contact the joist fabricator for info. The company name is normally on the tag.

 

[JedClampett]

I don't think that a W14 50 feet long is going to do anything. You're off the load table in AISC by 15 feet. As the earlier post said, there's a deflection already set in the span. Plus the stiffness of the existing joist, as flimsy as it is, is likely higher than the W14. If you deflect a weak spring the same as a weaker spring, he stiffer one takes the most load.
And can the interior W24 take the load increase? If so, I would shore up the roof, add 24 inch joists between the existing ones and double your capacity. You'll have to cut and re-attach your bridging. Your client can spend some of that money they saved using joists in the first place.

 

[dold]

First of all, 24" joists (H series or K series) span tables only go up to 48ft spans. A 24H8 at 48ft has an approximate moment of inertia of 328in^4 [based on K-series criteria of Ix = 26.767*W(LL)*(L-0.33)^3/1000000]. A W14x38 has Ix=340in^4.

You'll have to brace a W14x34 at the midpoint (at the very least) to meet a code check for 100plf load (assuming you've got about 50psf total load x 2ft o.c. = 100plf).

I don't like it. Not to mention the 1/2"x6" plate is doing pretty much nothing, except adding ~10plf dead load to the existing joists.

If you're planning on splicing the beam in the first place, you might consider using a deeper (W21?), cutting it wherever you need to in order to fit around the utilities, then cut the web out around the utilities, then splice it and reinforce the web as required.

 

[JRS56]

After much back and forth because of new building (PEMB) foundation requirements, it was decided to move the new building far enough away so drifting isn't a concern and the new foundation wouldn't interfere with existing foundation.