Added on roof - where trusses meet rafters

Have a look at the attached pictures and tell me what you think about it.
like the load transfer of bracing of the new roof to the old trusses etc.

Any input appreciated.

[ 042.jpg]( 042.jpg)
[FONT=Arial][ 043.jpg]( 043.jpg)[/FONT]
[ 044.jpg]( 044.jpg)

Thank you,



I would be asking for a set of prints please, needs to be confirmed on paper—:shock:

Where did you find this, out in County jurisdiction?

Nope, in the middle of the City of Phoenix. :ack!:



This looks like a house I inspected in Phoenix a month or two ago. House had some fire damage too. Wonder if it’s the same one.

That is unbelievable framing. I am simply stunned.

No this one did not have fire damage, but I had one of those fire damaged ones a couple of month ago. Here is a picture of the owners: I can do a roof better with 2x4s with a wider rafter spacing. But it was easy to guess that this one never pulled a permit.

[ 052.jpg]( 052.jpg)

How do you know there is any load transfer to the trusses through the “bracing”?

I know a good structural engineer here in Phoenix if you need to refer one.

I am not implying that the bracing was intended to transfere loads, however it pushes on the sheathing and the truss was not designed to handle the point loads on its web member. It might be minor but that is the reason I was hoping for opinions. I just know that some of the framing details at the connection points are missing.

Whom do you recommend Gary?

Whether there is a point load at all would depend on the strength and span of the addition rafters. If they are capable of spanning the entire distance between ridge and lower support, there may be little or no load transmitted through those braces to the trusses. If they are not, then of course, there could be a problem with the way the load is imposed on the trusses.

It may have just been a case of a carpenter with some extra wood up there that he didn’t want to carry back down…who knows? More all-encompassing pictures of the addition framing would be useful, but ultimately, the inspector should simply note unconventional framing and refer it to a qualified design professional, without offering an opinion of the structural correctness of what exists.


I am not sure I agree with this statement, but I think you know more about this stuff than me, so would you mind providing an explanation. If you provide an additional support at the mid span of a joist or rafter, the support will bear a load whether or not the joist/rafter was originally designed to carry the initial loads at it endpoints. The only way for this not to be the case is if there was zero deflection in the joist/rafter under load…which doesn’t happen. Am I right or Am I wrong?

You are theoretically right, but the load transmitted will be proportinal to the deflection, not necessarily half the load on the member. Framing members are designed for a full live load plus a dead load, and the full live load is rarely achieved in practice, although it would be more so on a roof in a snow area. In actual practice, in the pictured example, if the rafters were designed to span from ridge to wall, a very small load might be transmitted to the truss, under full live load conditions, but it is probably not enough to worry about. If the “designer” (if there was one) was counting on those struts to transfer load to the trusses, then yes, there is a big problem.

In a house where there are many partitions but only one load-bearing partition, the same possibility exists over all the non-load-bearing partitions.
Theoretically, second floor or ceiling joists would deflect and place a load of some sort on the non-load-bearing partitions, but do we design for that? We would spend more time calculating than drawing the plans if we did, and that is only if it were possible, and then we’d need all sorts of supports in the basement, or some mighty heavy first floor joists, to deal with all those loads.

Here in MI, that wouldnt last one heavy snowfall… something would give away.

There is not enough information given to support such a statement.