Cable-suspended roof

[danvines]

Hello all. I've been asked by a roofing contractor to assess an existing roof structure that is unlike anything I've ever seen.

The building was originally constructed as a high-school gymnasium in metro Atlanta in the late 1950s or early 1960s. It is oval in plan and appears to be a cast-in-place concrete frame with masonry and light-framed infill. There is a CIP concrete "compression" ring at the top which supports a grid of stranded cables (similar to unsheathed PT cables) at 3 to 4 feet o.c. each way. These cables support metal decking on which I understand there is a lightweight concrete or gypcrete topping of currently unknown thickness (assumed to be approximately 4"). There are NO significant connections between the decking and the cables. The decking appears to have simply been laid on top of the cables, although there are screws connecting the end and side laps of the decking sheets. I've attached a couple of photos.

Has anyone ever encountered (or better yet, designed) something like this? I've located and ordered a copy of a textbook, Cable-Suspended Roofs, by Prem Krishna (1978). I was hoping to be able to find someone with firsthand experience whose brain I could pick.

The goal is to be able to re-cover the existing 3-ply built-up roofing with a fully-adhered TPO membrane. I'm not really worried about the extra weight of that membrane, but there are several areas where the decking has corroded, and I'm unsure of the need to replace those areas. There is some surface rust on the cables, but from what I've been able to see/access, it doesn't look like it would affect their capacity. The method of attachment between the cables and the concrete compression ring is unknown, but there are no areas of obvious distress to the concrete at cable locations.

 

[Celt83]

We found some errors in that book a bit ago see this thread: https://www.eng-tips.com/threads/chain-catenary-vertical-reaction-help.463918/#post-8519755

 

[SWComposites]

" ... several areas where the decking has corroded, and I'm unsure of the need to replace those areas." meaning "rusted"? > are they planning to remove the decking? maybe remove the worst rusted deck panel and remove all the rust etc and measure remaining thickness. its likely the top surface is worse than the bottom.

"There is some surface rust on the cables, but from what I've been able to see/access, it doesn't look like it would affect their capacity" > I would be very careful with this as you are now going to be responsible for the cable strength. rust and paint should be removed and then carefully inspected.

 

[JAE]

The only type of roof I've seen myself is the Villita Assembly Hall in San Antonio, Texas.

This was a circular, cable roof assembly (not flat - a natural ponding concern!!) that was constructed of an outside compression ring and an inside tension ring, radial cables were pre-tensioned by pulling down on the interior tension ring, stretching the cables. Then concrete precast planks were placed on the cables. The tension ring was released, placing the concrete in compression, which held the concave shape you see in the photo.

This isn't much like your roof, though, as yours appears to simply use the cables as vertical support. Not sure how your roof handles uplift if the deck isn't actually attached.

 

[danvines]

The only type of roof I've seen myself is the Villita Assembly Hall in San Antonio, Texas.

This was a circular, cable roof assembly (not flat - a natural ponding concern!!) that was constructed of an outside compression ring and an inside tension ring, radial cables were pre-tensioned by pulling down on the interior tension ring, stretching the cables. Then concrete precast planks were placed on the cables. The tension ring was released, placing the concrete in compression, which held the concave shape you see in the photo.

This isn't much like your roof, though, as yours appears to simply use the cables as vertical support. Not sure how your roof handles uplift if the deck isn't actually attached.

The only possible mechanism for resisting uplift is the dead weight of however much “concrete” fill is on top of the metal decking.

We are still hopeful that original drawings can be located. Even if they are, we are probably going to take some cores to verify the thickness and properties/condition of whatever is there. Ensuring the crowing rig won’t be anywhere near a cable is going to be a trick.

Drainage on this roof really isn’t that bad. The saddle shape forces water to two drains near each low edge, and the very short parapet is likely adequate to accommodate emergency overflow. That said, we are considering adding a small amount of tapered insulation near the center of the roof to ensure positive drainage.

 

[danvines]

" ... several areas where the decking has corroded, and I'm unsure of the need to replace those areas." meaning "rusted"? > are they planning to remove the decking? maybe remove the worst rusted deck panel and remove all the rust etc and measure remaining thickness. its likely the top surface is worse than the bottom.

"There is some surface rust on the cables, but from what I've been able to see/access, it doesn't look like it would affect their capacity" > I would be very careful with this as you are now going to be responsible for the cable strength. rust and paint should be removed and then carefully inspected.

Ultimately, I hope to avoid replacing any decking (even where it has corroded through. I believe it was used to serve as a form deck for the concrete fill placed on top. To remove and replace it would be extremely difficult and would run the real risk of damaging the cables.

Removing paint to inspect the cables is definitely part of the plan, but at this time, I have no reason to suspect they are compromised.