Is it safe to remove the bottom chords of a series of garage roof trusses?

Question

See attached picture. Looks like an 8x10 cut-out. He wants to lift cars up so he can walk under them.
Is this safe? Doesn't removing the joists affect structural integrity?

Where are the bearing walls? It looks like the remaining ceiling joists could stabilize the roof structure, provided the bearing walls are far enough from the opening. — Lee Sam, Aug 19 '19 at 22:08
He just created this future situation for himself: https://diy.stackexchange.com/q/94143/42053 — MonkeyZeus, Aug 20 '19 at 16:14
With this done to the roof supports, I think his planned car lift might be even more scary. I hope it at least stands on the floor instead of being hung from the now even weaker roof beams. — jpa, Aug 21 '19 at 08:02
@jpa The car lift ground posts are clearly visible in the photo. — MonkeyZeus, Aug 21 '19 at 12:40
The next problem is probably the lift connection to the concrete. People assume their concrete floor is structural and hammer drill anchors into it, but it's a floating slab just waiting to break... Possibly without any rebar reinforcement. — Dotes, Aug 22 '19 at 13:08
Not pictured? Flying buttresses on the outside of the building.... or perhaps there are extremely fine cables (that don't show up well in the photo) are tying the sides of the roof rafters in place? — Neoheurist, Aug 22 '19 at 18:04
Did you notice how the roof angles out on the right side? Rafters attached to rafters. Shouldn't there be a load-bearing wall on the right where the "inner" rafters come down, like you can see on the left? Even with the horizontal joists still in place, that's a huge span for the bottom joist of a truss to support. I have a feeling our DIY specialist here not only cut the trusses but also did some haphazard removal of load-bearing walls before that to create the workshop he wanted. Time to call in a structural engineer! — Christoph, Aug 23 '19 at 12:45

isherwood · Answer 1 · 2019-08-21T13:37:35.303

68

Absolutely not safe.

Those trusses were engineered with a heavy (critical) dependency on the bottom chords, which are in tension. Removal has left them extremely vulnerable to collapse due to spreading, especially under snow loads, but also under just the load of the roof itself. The roof system is basically a hinge now.

To get a good mental image, picture yourself standing on an ice rink and sliding your feet outward. How far do you get before your groin starts to scream at you?

Best case scenario, that ridge slowly sags over the opening and looks ridiculous. It will push walls around with it, cracking drywall and binding windows. Worst case, full collapse. Death. Destruction. Worst of all, humiliation.

Repair will require new lumber installed with a suitably fastening system, such as construction adhesive and through-bolts. It may be necessary to winch the walls back into alignment first if sagging has already begun.

edited Aug 21 '19 at 13:37

answered Aug 20 '19 at 14:00

isherwood

119,766
7
148
349

1

There are a number of ways the damage could be repaired without lumber - but repaired it must be. – Martin Bonner supports Monica Aug 20 '19 at 14:06
10

Yes, but that probably goes beyond the DIY capabilities of anyone who would do this to their roof system. Also, any replacement for lumber must be laterally stabilized as those bottom chords were by the "rat run". – isherwood Aug 20 '19 at 14:07
It might be cheaper (and safer) to remove the roof, trusses and ceiling and install new trusses (vaulted type?) and new roof. – Keith McClary Aug 22 '19 at 00:08
1

To put the first sentence a different way: ***If those trusses were not necessary, the builder would not have put them there in the first place.*** Why spend more time and use more material, making something that has no purpose? – alephzero Aug 22 '19 at 20:41
@alephzero Because the builder wanted a ceiling. – Lee Sam Aug 27 '19 at 22:55

score 31 · Answer 2 · edited Aug 22 '19 at 14:42

31

With no bracing above, you might want to check the out side walls - they are probably already spreading. Once they start moving the stabilization and repair can cost many thousands if the roof stays in place, tens of thousands if it comes down.

There are ways to mitigate the damage done, but it needs to be done now before the walls spread, the rafters move and the roof leaks if not collapses. Yes I have seen bonehead DIY hacks like this.

Steel plates in the outside walls with cables to pull the walls back together, repair the damage to the rafters, new bracing with some engineered cables added.

I would suggest getting the bracing and cable system installed now or it will cost much more than the lift and the loss of use if the roof comes down.

edited Aug 22 '19 at 14:42

manassehkatz-Moving 2 Codidact

94,107
9
116
276

answered Aug 20 '19 at 07:53

Ed Beal

101,306
4
72
150

4

One thought, is that it *may* be possible to replace the joists with much beefier joists higher up. That would allow some of the additional headroom the owner wants, without compromising the structural integrity of the building. – Martin Bonner supports Monica Aug 20 '19 at 14:25
4

It may, but we have no way of knowing what the loads on those truss members are, and 2x4 top chords probably aren't adequate. You'd need to sister heavier boards on the complete A-frame. – isherwood Aug 20 '19 at 14:30
2

That’s the bracing I am talking about but depending on how high the car protrudes into the space will dictate if engineered cable supports are needed. (If the car sticks more than a foot or possibly 2 it will probably require an engineers stamp for whatever is done. Wish I could say I haven’t seen this kind of hack before. – Ed Beal Aug 20 '19 at 14:33

score 18 · Answer 3 · edited Aug 22 '19 at 11:42

18

The triangle is an extremely stable form of architecture precisely because it has three sides. You remove one side, and you have one of the least stable forms of architecture on account that two members connected at a point can be affected by torque, which is by definition a force multiplier. If you want to play around with it to get a sense, try gluing two Popsicle sticks and three Popsicle sticks like the design in the picture above.

Any force directed downward will because of the elasticity of the wood create force outwards. If the members were steel, there would be substantially less deformation resulting in less outward force, but lumber is relatively flexible.

Now, it should be noted that some of the horizontal members are still in place, as many as half, and those runs will make it unlikely the structure will collapse under normal conditions, however, heavy winds or heavy snowfall, as mentioned in other comments, likely to create outward stresses in the middle section of the structure. The exact details of what will happen will largely be a matter of detail. To some extent, existing construction code generally overengineers a solution to ensure that minor failure will not result in human harm.

It is possible to create both the space needed to elevate the car, and have a structurally sound garage given the picture. For instance, additional chords can be installed low enough to recreate trusses, but high enough to accommodate a raised vehicle. There are a number of truss styles that do not use a horizontal bottom chord, the scissor truss for instance. And using steel as a construction material can help to redirect the tensions through materials more capable of absorbing the load. Lastly, the walls themselves can be modified. It was typical of cathedrals to buttress on the exterior, the wall to handle high-arching ceilings, for instance.

Construction is often a question of engineering as much as it is design, and if safety is a concern, such as it might be in this photo, consulting a structural engineer would be the right thing to do. Structural design is their expertise, which would be my recommendation.

edited Aug 22 '19 at 11:42

Bob

103
4

answered Aug 20 '19 at 17:31

J D

583
2
9

For situations where one would need a large hole, would there be any approved ways of using horizontally-mounted beams or trusses to convey horizontal forces from the middle of the hole to the ends, or would every such design need to be independently professionally engineered? Given the roof pitch and amount of ceiling on either side of the hole, it would look like it should be easy to fit horizontal trusses that would be strong enough to handle such loading, – supercat Aug 20 '19 at 17:46
2

My intuition (by no means a calculation), would say that in this picture, several strategies would strengthen the roof. Running some sistered members across the cut joists would obviously create rectangles that would reinforce the wall from deforming. Purlins along the rafters would also work, or tying together those inside chords from the front to back parallel to the wall would also help. It's really a question of how sure one wants to be. Horizontal members higher up are a must. https://en.wikipedia.org/wiki/Timber_roof_truss#/media/File:Kehlbalkendach_mit_liegendem_Stuhl_1200px.png – J D Aug 20 '19 at 17:58
7

@supercat If you want a roof with a hole, you design a roof with a hole. You don't first design a roof without a hole, then cut a hole in it, and then try to fix up the resulting mess. – alephzero Aug 20 '19 at 18:04
1

@supercat This is the sort of timber roof design still standing after more than 600 years after it was built: https://www.greenoakcarpentry.co.uk/wp-content/uploads/Presentations/pageturners/AH%20Timber%20Joints%20Traditional%20and%20Contemporary/index.html#/9/zoomed – alephzero Aug 20 '19 at 18:16
I think in the design above, my personal choice would be to add to the sawed faces of the joists two sistered 2x6s, and then lay another 2x6 over the length to gusset those joints. I'd go with a purloin on the rafters tied to horizontal chords half way down, and then if I wanted to really strengthen the walls, lay a pratt truss horizontally on top of the joists near the top plates. – J D Aug 20 '19 at 18:42
1

@alephzero: Obviously one should design, engineer, and install reinforcement structures before removing cutting or otherwise removing structural members, rather than hoping that nothing collapses between the removal of old structure and the installation of reinforcement. The question would remain, though, whether there would be any commonly-approved methods to build a roof with a large hole whose edges don't have supports immediately under them. – supercat Aug 20 '19 at 19:47
3

Some good information in this answer, and some bad. It is absolutely nothing to do with timber vs steel; it is to do with *connections*, and also member size. – AndyT Aug 21 '19 at 09:38
1

Yes, it's foolish to speculate at repairs that meet the intent of the destructor without doing the math on the remaining truss members. No repair can succeed unless the components the repair is applied to are up to the task. Knowing what I do about modern engineered trusses, there's little room for experimentation. – isherwood Aug 21 '19 at 12:47
1

Repairs can succeed despite having mathematical certainty. Most repairs do not have mathematical certainty, but follow rules of thumb. To say it doesn't matter if a structure is made of timber with fasteners or welded steel isn't a true statement. – J D Aug 21 '19 at 14:35
As far as there are any commonly-approved methods for designing ceilings with holes in them, the answer to that is certainly yes. In architecture, the design pattern is called a vaulted ceiling. https://www.architecturelab.net/what-are-vaulted-ceilings/ – J D Aug 21 '19 at 15:03
There is nothing about the definition of torque which means it is a force multiplier. I don't think this answer adds anything to either of the earlier answers. – Martin Bonner supports Monica Aug 22 '19 at 12:38
@MartinBonner The definition of torque is T = F * r * sin(theta) which means that the further away the force is applied, the greater the torque since the force, by definition is literally multiplied in part by the distance. – J D Aug 22 '19 at 13:36
1

I agree with your definition of torque. I don't see how that means torque is "force multiplier". I think I know what you are trying to say, but i think you are failing to express it clearly. – Martin Bonner supports Monica Aug 22 '19 at 13:38
A force multiplier is anything which affects the magnitude of the force in an operation. For instance, when torquing a lag bolt into a joist, once might use a breaker bar of different lengths with the longer bar multiplying force by a greater amount. Hence, one can increase torque by using a longer length. The phrase force multiplier, besides its military sense, is used in physics when discussing machines. https://www.primaryconnections.org.au/sites/all/modules/primaryconnections/includes/SBR/data/Phy/sub/mechsys/mechsys.htm – J D Aug 22 '19 at 13:49
1

Even more than the lumber being flexible, the way *joints* are constructed in a wood truss are absolutely designed to NOT rotate. I think those would start to fail first. – StayOnTarget Aug 23 '19 at 11:32

Criggie · Answer 4 · 2019-08-23T23:57:35.510

As pictured - very bad idea.

It may work IF the roof is re-engineered to carry a lot more of the stress across the top and bottom of the cutout. That means a much beefier horizontal beam on all four sides of his hole AND a rework of the roof by re-trussing the entire structure.

Some examples:

This one is a Scissor Truss, and also wears the more general name "church truss" or "nave truss" Note some of these have thinner tie rods across the middle exactly where the bottom chord would be - those aren't suitable.

A little ornate for a workshop, but you can see how the load is being brought down to the walls with inner buttressing.

This one is called a Coffer Truss and as pictured would work perfectly. However the thickness of the rafters (top cords) would have to be increased several-fold to cope with the extra stresses. As pictured, this one would expect to have supporting walls/posts straight under the lower "corners" so would require a vertical post straight down beside the car lift's posts. That could be workable, if these new posts were properly footed into/onto the existing concrete floor.

This would work too - depending on how much height there is at the top of the wall.

Last option is to completely reframe the roof with steel trusses, and do away with the wooden ones. Properly enginerded steel will have more strength than the wooden ones while allowing the open space required.

Any way - whatever creative handyman did this has made a difficult job into an utter nightmare of a job. It would have been less work to remove the roof and then re-truss leaving the bottom chords in place to support the walls until the new roof framing is on.

Frankly, it would probably have been cheaper to build a new garage rather than monster the existing one and now have to repair it.

Also consider any insurance claim will be harder to win, based on the level of care exhibited here. I'd bet any insurance claim for snow damage or storm damage would be rejected out of hand. If this is used for a business, any claim for a customer car damaged by this collapse would be rejected as "failure to use due care"

You need to speak with a structural engineer immediately in person and ideally on-site. I am not a structural engineer.

The coffer truss as depicted is not suitable in this situation. The image indicates that there are four load-bearing points: The two outside walls but also two more points further in where the outer triangles end. Without those two additional load-bearing points, little would change in the shearing force of the roof against the outside walls. Only difference would be that the roof collapses in three pieces. — Christoph, Aug 23 '19 at 12:36

Lee Sam · Answer 5 · 2019-08-29T11:45:31.950

0

I think it’s safe, if...

Now that everyone has upvoted this as unsafe, I’ve got to take the other side of the argument for several reasons: 1) it’s not a truss, 2) it’s “stick-built”, 3) either side could be removed and the remainder would be stable, 4) horizontal structural support is elsewhere other than in the center horizontal span,

1) Look close and you’ll see that it’s not a truss. The structure does not create a truss spanning the room...or it would have collapsed immediately...especially with that much removed.

2) The members are stick built. The roof member spans from ridge to next support...diagonal brace. The roof member is in both tension and compression. A truss member is in EITHER tension or compression. (That’s why “double bending” in truss design is so dangerous.) The loads are transferred down the diagonal members to the walls or cantilevered beams over walls. In truss construction, loads are transferred to the END of the truss.

3) If you look at the right side you’ll see that it could be removed and the left side would remain structurally sound. That’s an indication that it’s not a truss structure , but a stick built structure. (One half of the roof structure looks like a picnic shelter.)

4) The horizontal stability is in the end walls, NOT the center ceiling support. That’s why you don’t see large cracks in the ceiling gypsum board.

Now, I said “if” at the beginning. The “if” is in reference to the need that each side must be horizontally stable and has at least 2 bearing points on each side of the opening.....and obviously it is so far.

edited Aug 29 '19 at 11:45

answered Aug 25 '19 at 22:30

Lee Sam

22,822
3
19
42

1

They sure look like engineered trusses to me. The photo is lousy, but I believe I see at least one steel gusset at the peak. What superficially appears to be a ridge board is more likely a gap in the sheathing for ridge venting. Also, hand-framed roofs almost always had 2x6 or larger rafters for spans that long. If you have more concrete information, please share that. – isherwood Aug 26 '19 at 14:43
As to the immediate collapse... no. The sheathing is acting as a significant shear brace right now, and the bearing walls may have buttressing from intersecting walls or roof structures. – isherwood Aug 26 '19 at 14:45
@isherwood Nope, it’s not a truss and there is no gusset plate. Look at the diagonal member where it intersects the horizontal member...it laps the horizontal member and there’s no gusset plate. – Lee Sam Aug 26 '19 at 15:56
@isherwood It hasn’t collapsed immediately because it’s not a truss. The loads are transferred “down” the walls, not “out” to the outside walls...that’s why it hasn’t collapsed. – Lee Sam Aug 26 '19 at 16:03
I think you may be right. I missed that clue. There's merit in your answer in that case, though the layperson is probably not wise to do what this guy did with so many variables at play. – isherwood Aug 27 '19 at 13:13
@isherwood I wish the original OP would give us an update. – Lee Sam Aug 27 '19 at 16:08
@leesam The intersection of the horizontal member and diagonal member is not visible in the picture. That "lap" is the horizontal running in front of the diagonal of the next truss back. If this were a cut and stack roof, that diagonal would be supported by a structural member like a load bearing wall or beam. I vote truss. – Valkor Aug 28 '19 at 20:03
@Valkor The support does not need to be directly under the supporting load. I’m not saying it was designed this way, but often we’ll design a cantilever to support loads. In this case, they were lucky that they cut the horizontal member in a location where a cantilever would support the load. So, the horizontal member is now a cantilever that happens to support the roof load. Can the horizontal member structurally support the roof load??? We won’t know until we know the sizes of the members and their spans, but it’s not (nor has it ever been) a truss. – Lee Sam Aug 28 '19 at 20:18

Is it safe to remove the bottom chords of a series of garage roof trusses?

5 Answers5

Linked