Masonry Fire walls 1

[delt]

We have a project that includes a fire separation wall. As per architects, the wall construction is this: Each side of the wall must be designed as two separate structures. This is not a problem. The issue is that there is a large bank of doors in this wall. Therefore, a third separation wall is to be placed in between the two other walls, which now have become steel beam and columns in lieu of bearing walls. This middle fire wall needs to remain standing if one or the other sides falls. This means that the wall needs to be designed as a cantilevered wall without accounting for lateral support from either wall.

Does anyone have any experience with such a situation? What have you done in the past? Is there something I am missing here? A 30 feet high cantilevered wall exposed to wind loading requires an unbelievable amount of reinforcing, if possible at all!

Any assistance is appreciated.

 

[JAE]

Generally you would use a double wall, each capable of qualifying for the necessary fire resistance and each connected to the opposite structure...with a small gap between...perhaps 2 inches.

 

[spats]

You only need one wall, and it doesn't have to be freestanding. If I understand you correctly, you now have steel framing on both sides of the firewall. You can attach the wall to the steel framing on both sides with "break-away" connections. These connections need to break loose on the fire side of the wall in the event of a fire. When the structural members on the fire side sag under the heat, the end rotation of the members causes them to slip off of their connections (if detailed properly). Factory Mutual has a lot of good info on this, including the sag forces and a means of calculating the end rotation of the members. Factory Mutual calls these MFL (maximum forseeable loss) firwalls.

I'm not sure I understand why a wall in the middle of steel framing (normally interior) would be exposed to wind, but in any event, the wall and it's foundation can be much lighter than what you would have with a cantilever.

 

[delt]

Thanks for the info. The reason we can't have two separate bearing walls is that the architects have a problem with doors in both walls - this creates an ingress/egress problem. The middle wall will be the separation and have the doors.

The reason the middle separation wall would be exposed to wind is that when one structure falls, the middle wall is now an exterior wall, exposed to windward pressures.

Thanks for the information on Factory Mutual - I have heard Hilti may have some help as well.

 

[spats]

I would think that if a portion of the building burned down, it would be considered a temporary condition. They likely would not maintain it in that condition for an extended period of time. Designing for a 50-year wind event seems a little extreme.

I'm sure there are other sources, but the Tilt-Up Concrete Association has guidelines for "construction period" winds. They basically reccommend 70 mph for 135 mph regions and less.

 

[JAE]

spats - does FM have details of this break away connection available somewhere? We've tried in the past to develop something like that but could not find this info. Got anything on this?

Thanks,

 

[delt]

I found this page - interesting info. on firewalls, but nothing on "break-away" bolts. Neither does Hilti. Any info. anyone has on such an assembly would be extremely appreciated. The link to the page is:

http://www.beijingnews.com/courses/g130/g130.htm

In discussions with our architects, their is a UL assembly for a similar situation with "break-away" connections to each wall in apartment construction (wood frame). The thought is that each wall would provide lateral stability, with a clip angle attaching the firewall to each adjacent wall.

However, we can find no such connection/approved assembly for masonry walls (because masonry walls typically don't need such an approved assembly).

Does anyone know of a possible clip that would provide lateral resistance, but in case of failure, would fail before pulling the middle fire wall with it?

Again, any help is appreciated.

 

[Taro]

Get a copy of NFPA 221 Standard for Fire Walls and Fire Barrier Walls. It addresses the various types of walls you can use (cantilevered/freestanding, tied, double).

 

[spats]

There are no "break-away" bolts, only fasteners designed for explosion venting, but I wouldn't use them.

There are several ways to accomplish break-away connections. Following are two senarios and my solutions:

1. Beam/column framing along wall on both sides: I would hard connect the top of the columns to the wall with clip angles, and bottom bear the beams on column cap plates. Weld unheaded studs to the cap plate installed throuth horizontal slots in the bottom flange of the beam. The length and position of the slots is such that when the beam on the fire side sags under heat, the end will rotate off of the studs. The studs need to be short enough to clear the back of the slot in the beam flange so it can so it can pull free when rotated. The fire side will exert a sag force on the connections on the opposite side of the wall, which they must be able to safely resist. If the connections on the non-fire side are the same connection, the pull would deflect the wall over slightly until the stud reaches the end of the non-fire beam slot, and the wall will be safely held.

Because of roof uplift, it is also necessary to use something like Z-clips welded to the column cap that lap over, unattached, to each edge of the beam flange. They should be placed near the point of rotation so they don't restrict end rotation of the beam.

Diaphragm connections for forces along the wall can be made with the same unheaded studs through oversized holes or short slots in the roof edge angle, with no in/out connection.The edge angle can even be held a little off the wall so it can more easily bend the studs and slip off if the edge angle wants to go with the rest of the falling steel. If possible, I would try to avoid using the firewall as a shearwall, because these "break-away" connections become a little more problematic (unless you have a better idea).

2.Edge joist parallel to the wall: connect a horizontal plate with an short upturned lip to the wall that hooks under and around the top chord of the joist, at intermittent locations along the wall. The clip is light enough (gauge material) that if the roof structure on the fire side starts to sag under heat, the clips will deflect and the joist will pull free. The same clip on the opposite side prevents the wall from being pulled over with the collapse (the clip is strong in tension).

If both of these conditions occured simultaneously, you could probably even justify eliminating the studs at the beams (Z-clips only).

There is no reason why you couldn't use similar concepts while even using the wall as a bearing wall. Joists could be connected similar to what I described for beams, etc.

I can't say that these concepts have necessarily been fire tested, but the concepts are sound and are accepted by the fire insurers, in my experience. I'm sure you can probably come up with some of your own concepts.

 

[Stress02]

I have done a multi-story cantilever wall. Turned out that concrete was an easier task than masonry.

 

[jike]

Heckmann Building Products manufactures Break Away Firewall Anchors in various materials that melt at temperatures over 800 degrees F. These anchors can be used for stabilizing multi-story firewalls.

See the following URL.

http://www.heckmannbuildingprods.com/PFire.htm

You can talk to one of their Tech Reps for more info.

 

[delt]

I noticed the Heckmann products, too. I discussed this with a Hohmann and Barnard Rep. - they can do the same thing. Basically, they manufacture most of their anchors out of a metal with a lower melting point than steel. My question is - how do you guarantee the anchor will get to this temperature before failure? The fire causing the failure could be away from the fire wall. This fire could still cause a failure, but the anchors in question would not be melted, causing probable failure of the fire wall. Any thoughts?

 

[jike]

delt:

You have expressed a concern that I have always wondered about.

I would guess that if the fire is away from the firewall and there is a partial collapse (away from the firewall), that the remaining structure and unmelted anchors would generally be sufficient to prevent the collapse of the firewall.

If the fire is adjacent to the firewall, the anchors would melt on that side of the wall and the unmelted anchors and structure on the opposite side would provide stability to the firewall.

 

[spats]

The actual scenario would be a fire in the bay immediately adjacent to the firewall (consider it in the middle of the bay) which would cause the steel to sag and fail, with bolts immediately adjacent to the wall being largely uneffected by the heat. As I said earlier, I would forget about connections that actually break. See my January 23rd reply.

 

[sc]

This is a classic case of a picture saying a thousand words . . .

The steel framing design to either side of the wall only needs to meet the fire rating design period. What happens after this is not as important as ensuring sufficient time for occupants to get out. Also if the fire wall does its job, then only one side will be subjected to failure whilst the other side supplies stability.

If you have doors in the fire wall, they need to be fire rated and self closing (also incapable of being propped open). If the doors don't meet the fire rating criteria, then you do not have a fire wall. Generally it is better practice not to have any openings in the fire wall.

AS4100 specifies design criteria for steel that needs to be fire rated.

regards
sc

 

[jike]

Spats:

I was suggesting the use of break away connections for the multi-story applications, not the single story roof applications.

 

[spats]

With regard to SC's comments: Insurance companies like to limit their losses, not just save lives. That is why it is necessary to go beyond simply insuring that people have time to get out. Factory Mutual calls these MFL (maximum foreseeable loss) walls, and the wall must stay even if there is a failure on one side of the wall, so the fire doesn't spread.

 

[jheidt2543]

I talked with Heckman Building Products Technical Services today (3/15/04) regarding this topic. They do make the "melt away" anchors, however none of them have been approved as yet by any of the major code authorities or by UL. I was told that Heckman is developing the submittals, but the testing and approvals have not been done yet. They have sold the units to a number of projects though.

I also called my local state plan approval department and they basically said they will review and CONSIDER approval on a case by case basis only. They have not given a blanket approval to any one design detail, with or without the "melt away" anchors. This sounds pretty iffy to me and could entail a LOT of work.

Have any of you actually used these things? I'm sure that there are other ways to handle this, but I haven't found any yet, how about use'guys? Have you seen or used any details that work?

 

[spats]

To me this is a classic case of somebody thinking they're going to invent a better mousetrap and corner the market, without thinking how impractical the solution is. You cannot reliably predict that "melt-away" anchors would perform as intended. As I said earlier, if the fire is away from the wall the anchors would not melt, but the sagging and collapsing structure could easily exert enough force to pull the wall with it.

I have successfully, from an approval standpoint, used connections similar to what I described in my January 23rd response above on at least a half-a-dozen projects.

 

[jike]

How do you detail 3 or 4 story firewalls? What about if the floors and roofs are not at the same elevation?