Cutting attic trusses

[Pawel1234]

Hi All,

I am new to the forum and this is my first post. I am looking forward to contributing and broadening my engineering experience.

I would like to know your opinion on how to best approach truss modifications where existing attic trusses need to be cut to form openings, for example, for velux windows or dormers.

Would you support the cut trusses only at velux bridles, as per sketch a (please see attached sketch(, at velux bridles and a ridge beam/top hat beams (sketch b) or at all node points (skecth c)?

Obviously from a contractor's point of view a smaller number of supports would be the best option. On the other hand, the addtional supports will provide better support to the cut trusses, potentially limiting any increased deflections resulting from the loss of triangulation after cutting. Also, the addtitonal supports will distribute load from the cut truss more evenly to adjacent doubled/tripled trusses potentially making the whole arrangement more structurally efficient.

Thanks in advance.

 

[r13]

What size is the opening, and spacing of the roof truss? Is there an option not to cut the truss?

 

[oldestguy]

Those are not trusses. Rafters.

 

[r13]

Woops. Yes, rafter, or top chord.

 

[BAretired]

Welcome to Eng-Tips, Pawel1234.

Snow loads vary. Truss slopes vary. Each situation has to be reviewed separately. There are no general rules for such an operation. Just ensure that you can satisfy statics.

Option a) works well enough if the cut rafter is adequate to prevent excessive sag of ceiling joists which act as rafter ties.
Option b) is likely only required when the velux windows are unusually wide.
Option c) minus the ridge beam might be required for ceiling support, but I would drop the beams to the bottom of the verticals for ease of fastening.


BA

 

[dhengr]

Pawel1234:
For removing one or two rafters, spaced at 16 or 24”, of reasonable length and with normal roof loads, most good carpenters or engineers would just header across at the top and bottom of the opening, per your sketch a). They would also double the adjacent existing rafters in this process, in effect, to replace the roof load cap’y. of the cut rafter or two. That will usually pass a more rigorous engineering analysis. Pay some attention to the increased bearing loads (reactions) at the doubled adjacent rafters, at the ext. wall bearing points and at the ridge board, and supporting the roof sheathing (roof diaphragm). Under normal conditions removing more than two rafters starts to req’r. some serious engineering consideration.

If those are truly prefabed trusses, the added blocking or bracing shown in sketches b) & c) is probably appropriate to improve the load sharing situation, and you should double the remaining adjacent top chord members. When you cut rafters or trusses like that, you kinda screw up the lateral thrust picture at the ext. wall bearing points and at the ridge, and you should account for that lateral load change. The ridge is of particular concern because the added 2x chord members have nothing to bear against, across the ridge. In the extreme, the uncut roof plane becomes a lean-to roof surface without anything to react against, at the ridge, when the new opening becomes too wide, without some special attention to this issue.

 

[KootK]

1) I'm going to assume that this is not a metal plate connected attic truss as the term is typically used in north america. If it is, however, do let me know as I've a good deal of design experience with those. They're my second least favorite kind of prefabricated truss.

Also, the addtitonal supports will distribute load from the cut truss more evenly to adjacent doubled/tripled trusses potentially making the whole arrangement more structurally efficient.

2) Do doubled/tripled frames already exist or or are you proposing to add them?

3) What's the maximum number of trusses that you anticipate needing to cut? For now, I'm assuming that you'll only need to cut one per window.

4) My preferred solution to something like this would be similar to your option [a] but done with any eye towards restoring the original, primary triangulation:

a) Install headers as suggested to deal with what I've called the "micro" effects shown below. Local bending and shear in the members.

b) Give the headers from [a] a meaningful down slope stiffness and a path for compression load transfer through the system. The idea is to attempt to restore that all important, three member, primary truss action that you and others mentioned above. In my opinion, it is that mechanism that really makes these framing schemes successful and it should be reconstituted to as great a degree as possible. The sheathing alone will accomplish this to a degree but I prefer to augment that load path with something more discrete and readily evaluated as shown below. To that, you'd add any additional reinforcement you determined necessary.

 

[Pawel1234]

Thank you for your replies. You have raised some very interesting points. I hope you are all well. I am sorry I could not follow on your posts earlier.

Please see below a sketch of how I see the behaviour of a cut truss if it is only supported at proposed opening by headers.

I understand that due to cutting and re-supporting the rafters of the attic trusses at the headers the original rafter continuity is lost and pins are ‘created’ at header-rafter connection. This leads to the cut truss deflecting more than the adjacent trusses, possibly leading to issues with differential deflection.

At the same time, the double/triple trusses supporting the cut trusses will receive 2 point loads from headers, which due to asymmetric nature of the loading, will potentially cause the supporting double/triple trusses to deflect substantially.

Perhaps I am missing something and actually the behaviour is not as ‘adverse’ – what do you think?

My other concern is also the capacity of connections between the truss members in both the cut truss and in the trusses supporting the cut trusses due to different arrangement of loads and/or larger loads. How do you normally address the connection issue?

Hence, to cater for both the deflection and connection capacity aspects I would understand that both the cut truss should be supported at as many points as possible and all of the members in the supporting truss should be doubled. However, this obviously involves a lot of work. Do you see the same problems here or perhaps I am missing some points?

retired13:
It is more of a general question, but the opening width typically required is around 31” and the truss spacing is around 24’’. Therefore, at least one truss needs to be cut. I agree - it would be the best option not to cut trusses, however, generally wider openings are preferred.

---------

dhengr:
Could you please expand on the considerations you mentioned as per the below?

‘Under normal conditions removing more than two rafters starts to req’r. some serious engineering consideration.’

Would you say that in the case of the prefabricated trusses (and possibly trusses made of cut timbers on site) it is justifiable to double/triple all truss members to ensure that the load transfer between the truss members is that as in the original truss and increased forces in connections are catered for (assuming that doubling the trusses you also ‘double’ all of the connections?

With regards to the below, could you please illustrate/explain the issue with the ‘lean-to surface’ and how would you address it?

‘In the extreme, the uncut roof plane becomes a lean-to roof surface without anything to react against, at the ridge, when the new opening becomes too wide, without some special attention to this issue.’

---------

KootK:
1) I am looking for a general picture, therefore, my query applies to both prefabricated – metal plate connected attic trusses and also to trusses made of cut timbers on site with ridge board etc. From your experience, what are most important points to consider when cutting attic trusses made of metal frame connected trusses to form openings?

2) I am proposing to double/triple the existing trusses.

3) I would typically need to cut one truss, but sometimes two.

4) That’s an interesting idea in that there would be no increase in compression force in the rafters supporting the cut truss and thus no increase in the bottom chord (ceiling tie) tension. Therefore, the need to reinforce the connections in the supporting trusses could be avoided – in particular - rafter-bottom chord connection.

However, perhaps I am missing some point, but It would appear to me that the compression force from the cut truss rafter above would be transferred directly to the supporting truss rafter than to the rafter of the cut truss below. Wouldn’t some kind of slotted connection be required to achieve this?

 

[XR250]

@Koot,

Seems like you would need to install a similar mechanism in the ceiling plane for this to be effective.

 

[KootK]

Seems like you would need to install a similar mechanism in the ceiling plane for this to be effective.

I don't see it. Can you elaborate on your concern? You know, if you're interested...

 

[KootK]

From your experience, what are most important points to consider when cutting attic trusses made of metal frame connected trusses to form openings?

Metal plate connected attic trusses are a strange hybrid of truss and moment frame. The design of them is quite sophisticated and heavily influenced by unbalanced loading scenarios.
Frankly, I doubt that any of the suppliers could actually do it properly without the aid of the software that they use. I certainly couldn't have back when I was doing that work.
Similarly, I'd caution most conventionally trained EOR's against trying to design modifications to such trusses unless they could somehow create a situation in which they were confident that the metal connector plates were not taxed any more heavily than they were in the original design.

In short, for me, metal plate connected attics = abandon the cut trusses and rely heavily on beefed up members at the sides of the openings. Of course, slipping a new truss into an existing attic plenum isn't exactly the easiest thing either.

Wouldn’t some kind of slotted connection be required to achieve this?

Whether it would be required or not, I don't believe that there's a practical way to accomplish such a thing.

However, perhaps I am missing some point, but It would appear to me that the compression force from the cut truss rafter above would be transferred directly to the supporting truss rafter than to the rafter of the cut truss below.

I don't see that as in insurmountable issue. I feel that the compression will be introduced into the adjacent truss rafter for only the height of the opening. After that, it would get transferred back into the the other segment of the cut rafter. You'd also be naturally introducing a reinforcing bit across the height of the opening (my sketch again) that would share in resisting that compression load as required. The setup is heavily depending on the stiffness of the new horizontal pieces of course.

 

[Pawel1234]

@KootK,

In short, for me, metal plate connected attics = abandon the cut trusses and rely heavily on beefed up members at the sides of the openings. Of course, slipping a new truss into an existing attic plenum isn't exactly the easiest thing either.

I would think the same. However, I think this could also apply to any other type of truss, for example, trusses made on site with members simply lapped and nailed together which most likely were not designed at all - as is the case for many roof structures built in the past.

Would you mean to simply design/specify a new truss, possibly with two or three plies, along the exisitng trusses to carry the beams supporting cut trusses? The alternative of this could be possibly some cranked steel beam or some form of a steel raised tie-truss?

Having said the above, most clients/architects would likely think that it would be excessive to construct such a new structure just to form a window opening. Considering other types of trusses than the metal plate connected ones, wouldn't it be however more cost effective to introduce a new structure rather than to reinforce the existing trusses that need to support the cut trusses? I would understand that reinforcing the existing truss would first of all consume a lot of engineering time for site visits - for example to check the existing connections etc. with a view of reinforcing, isn't it?