Attachment of thermal collectors to decking or joists

[OHEcoEng]

Hello, all. I am currently working on establishing a connection method for solar thermal collectors to a roof deck.

We're good with the additional dead load, but I am concerned with the proposed attachment method.

Background:
The system consists of two parallel timber "curbs" 6"x12"x 24' long, each weighing approximately 450 lb and spaced at approximately 5.1' apart. The collectors are mounted at a 45 deg. angle between these curbs, and weigh 985 lb when filled with glycol. The curbs sit at a 45 degree angle to 24' 18LH04 joists.

With our 90mph wind speed using ASCE 7-05 horizontal wind pressures, I am looking at around 1900 lb of additional uplift alng the rear curb when considering rotation around the front curb.

The previous engineer approved of simply placing allthread through the timbers, through the 3" deep-rib roof deck, and through some unistrut against the bottom of the decking with bolt and washer. I don't like the idea of a few spot welds resisting this force.

What are your thoughts on this? Anyone know of a (better) established attachment method?

I have spoken with this "structural" engineer once before, and he was very puzzled about what ASCE 7-05 was. I seriously doubt that he gave the attachment method much thought.

The jpg shows how this thing is laid out - decking/insulation is not shown.

Mike

 

[JAE]

Here are the things I would do:

1. Add horizontal cross members between joists (connected to the joist top chords) in line with the through-bolt connectors. These could be tubes, angles, etc. welded to full depth vertical angles that are welded to the top and bottom joist chords.

2. Have the through-bolts thread through these horizontal cross members, not the deck as you properly want to avoid.

2. Add bridging (X) between the joists to further distribute load.

3. Check the joists for the added dead and wind loading. Strengthen the joists if necessary. If strengthening isn't workable, add new joists or steel beams between the existing joists.

4. Check any supporting beams that support the joists. Check the columns as well.

5. If you are in snow country, include snow drifting as appropriate.

6. If you are in a generous (and scrappy) mood - send the engineer a brand new copy of ASCE 7 with your compliments.

7. If you are in a darker mood - contact your local engineering board and warn them of this engineer.

 

[OHEcoEng]

Thank you, JAE. Advice taken. I know I'm going to run into resistance on this, especially when it comes to welding.

The good news is that it is apparent (after reviewing the joist capacities compared to what is there) that this section of the building was designed to supports some hefty equipment. There are also multiple stand-off in place in this area with nothing on them.

Good to hear some confirmation regarding my suspicions...

 

[JAE]

Hey - one more little tid-bit.

When you add a concentrated load to a joist, it is typically recommended that the load, somehow, get transferred to a joist panel point. This would require a small angle extending from your cross member down at some variable angle to attach near a diagonal intersection at the bottom chord. If the cross member actually aligns with a panel point (say within 4") then this isn't requuired.

 

[ToadJones]

what does "ASCE" stand for?