10,028 views ·
41 replies
10k views
41 replies
Glue nailing on existing beams or distribute on cc30 instead? Which gives the best result?
Page 1 of 3
I plan to stiffen up the floor structure a bit.
Currently, I have approximately 70x170 beams with a cc60 spacing. These are old beams, so some are not entirely square, meaning they're almost "half-moons," some with bark still on them, and the span is about 3.6 m.
When walking on the floor, you can really feel the vibrations, and I want to reduce them as much as possible.
Now I was thinking of adding at least K24 (preferably K30 or K35) 45x170 to stiffen it up and glue-nail them perfectly level so I get a good foundation for the floor afterward.
But now the question is the following:
Will I get the best result by glue-nailing the new beams to the old ones, so the new ones stiffen up the old ones?
Or is it better if I instead place a new 45x170 right in the middle of each section so I get cc30 spacing instead? (this method will be twice as difficult to crossbrace, not to mention 3-4 times as difficult because it will be cramped)
I imagine there will be slightly different effects with the different methods, i.e., both pros and cons?
Ideally, doing both would be best anyway.
Currently, I have approximately 70x170 beams with a cc60 spacing. These are old beams, so some are not entirely square, meaning they're almost "half-moons," some with bark still on them, and the span is about 3.6 m.
When walking on the floor, you can really feel the vibrations, and I want to reduce them as much as possible.
Now I was thinking of adding at least K24 (preferably K30 or K35) 45x170 to stiffen it up and glue-nail them perfectly level so I get a good foundation for the floor afterward.
But now the question is the following:
Will I get the best result by glue-nailing the new beams to the old ones, so the new ones stiffen up the old ones?
Or is it better if I instead place a new 45x170 right in the middle of each section so I get cc30 spacing instead? (this method will be twice as difficult to crossbrace, not to mention 3-4 times as difficult because it will be cramped)
I imagine there will be slightly different effects with the different methods, i.e., both pros and cons?
Ideally, doing both would be best anyway.
Last edited:
Interesting. I'm trying to understand the physics behind it all. What makes it much better with cc30 instead of glue nailing on existing beams? So what are the advantages of that, and what are the disadvantages of glue nailing them to the existing beams?huggan said:
What speaks for cc30, among other things, is that the floorboards themselves become a bit more stable between the old beams. I wonder how it affects the springiness as a whole though.
The width of a regel does not make much of a difference. Therefore, reglar are not wider than 45mm today. If you attach an additional regel to the side, there will be almost no difference. However, if you place new reglar between each regel, it becomes a completely different case. The floor flexes more the greater the distance between the reglar.
170 rule at 3.6m span. Am I the only one who thinks it sounds a bit weak? cc30 is really difficult to work with as it gets awfully cramped. Personally, I wouldn't even have considered anything else for that span. Is it not possible to get some support in the middle?
According to the Träguiden tables, 220mm is required at cc60, so cc30 is definitely required here. However, you can screw-glue the floor chipboard instead of using cross-bracing. If you only have floorboards, you will need to use cross-bracing.
Here came the interesting posts. And I have some answers to these.
After all, it ends up being just as much wood whether you install double beams cc60 as when you install single beams cc30.
What I believe makes the most difference is the nail-gluing, that is, the glue and nails, and potentially a few other things, which is why I'm asking in the first place. I'll get to that a little further down here.

I myself think it’s weak with 70x170 as mentioned, but then let me show you the Swedish Forest Industries' dimensioning table here.
If you need to use 170mm height on the floor joists, you can use the following according to the table:
Best to worst:
45x170 (C24), cc30, glued chipboard - max free span 4.10 m
45x170 (C24), cc40, glued chipboard - max free span 3.75 m
45x170 (C14), cc30, glued chipboard - max free span 3.55 m
45x170 (C24), cc30, nailed chipboard or tongue and groove paneling - max free span 3.18 m
Surprisingly, you can have over 4m if you just have cc30 and glue the chipboards? Or have I read the table completely wrong?
Unfortunately, I’m approaching the last option, since I'll have side-screwed 28x120mm tongue and groove planks (spruce).
However, I can always use K30 or K35, depending on how much that improves things, and also use cross-nailing/blocking.
PDF, see table on page 10
Source: http://www.svenskttra.se/publikationer/lathunden
Okay, the reason it all gets a bit complicated and I'm asking the original question is that the existing beams are very different in profile. Some are very nice square 70x170, others are basically half-moons where only the middle of them is 70x170, and up top and down below, they may only be 20-30 mm wide, which in turn weakens them. They are simply not as tightly controlled as today's wood. They used what they had at the time (it's nearly a century since the house was built).
At the same time as they aren't perfect 70x170, several of them have probably hardened significantly over time, which should in turn stiffen up a bit and give some extra stability against flex?
So the question really remains if it is more worthwhile to bind new K24, K30, or K35 with the old beams through nail-gluing or to let the new ones stand alone in the middle of each section. For the floor flexing between the actual beams, it probably doesn’t matter since the floor thickness will be 28mm (maybe 25 at worst after nailing), which is sufficient regardless.
In a way, it feels better to distribute cc30, so it doesn't become so much layer on layer, while on the other hand, it feels better to stiffen up existing "weak" "half-moon" beams so they don't just "flap" around there
Again, it would probably be best to do both, and then it would basically be like setting cc20
Well, isn't it the case that single cc30, as opposed to double cc60, only improves flex between each section? That is, the flex of the floorboards/flooring that sits on top? While the floor as a whole will flex roughly the same amount as if you put double beams in width?huggan said:The width of a beam doesn't matter that much. That's why beams aren't wider than 45mm today. Adding another beam to the side makes almost no difference. However, if you place new beams between each current one, it makes a significant difference. The floor flexes more the longer the distance between the beams.
After all, it ends up being just as much wood whether you install double beams cc60 as when you install single beams cc30.
What I believe makes the most difference is the nail-gluing, that is, the glue and nails, and potentially a few other things, which is why I'm asking in the first place. I'll get to that a little further down here.
I also think it sounds weak, but the house is from the 1920s, so I assume the standards were a bit different then. I've seen worse than this, I can say, for example, 120x120 at spans up to 4 meters. That is not funjohel572 said:
I myself think it’s weak with 70x170 as mentioned, but then let me show you the Swedish Forest Industries' dimensioning table here.
If you need to use 170mm height on the floor joists, you can use the following according to the table:
Best to worst:
45x170 (C24), cc30, glued chipboard - max free span 4.10 m
45x170 (C24), cc40, glued chipboard - max free span 3.75 m
45x170 (C14), cc30, glued chipboard - max free span 3.55 m
45x170 (C24), cc30, nailed chipboard or tongue and groove paneling - max free span 3.18 m
Surprisingly, you can have over 4m if you just have cc30 and glue the chipboards? Or have I read the table completely wrong?
Unfortunately, I’m approaching the last option, since I'll have side-screwed 28x120mm tongue and groove planks (spruce).
However, I can always use K30 or K35, depending on how much that improves things, and also use cross-nailing/blocking.
PDF, see table on page 10
Source: http://www.svenskttra.se/publikationer/lathunden
I think a bit like this too, just need to find facts about it.sblixten said:
I can’t use chipboards since we want a solid wood floor and at the same time, we don't want the floor to be higher in the room I'm now reinforcing, as the passage between the rooms is meant to be without thresholds/level differences.Forbo said:
Okay, the reason it all gets a bit complicated and I'm asking the original question is that the existing beams are very different in profile. Some are very nice square 70x170, others are basically half-moons where only the middle of them is 70x170, and up top and down below, they may only be 20-30 mm wide, which in turn weakens them. They are simply not as tightly controlled as today's wood. They used what they had at the time (it's nearly a century since the house was built).
At the same time as they aren't perfect 70x170, several of them have probably hardened significantly over time, which should in turn stiffen up a bit and give some extra stability against flex?
So the question really remains if it is more worthwhile to bind new K24, K30, or K35 with the old beams through nail-gluing or to let the new ones stand alone in the middle of each section. For the floor flexing between the actual beams, it probably doesn’t matter since the floor thickness will be 28mm (maybe 25 at worst after nailing), which is sufficient regardless.
In a way, it feels better to distribute cc30, so it doesn't become so much layer on layer, while on the other hand, it feels better to stiffen up existing "weak" "half-moon" beams so they don't just "flap" around there
Again, it would probably be best to do both, and then it would basically be like setting cc20
Last edited:
If you want to reinforce a floor, there are three options that are commonly used. Either from cc 600 mm to 300 mm, apply self-leveling compound on top of the floor, or install floor gypsum. Read all the rules related to tiling a wooden joist structure. These are the recommended methods.
Last edited:
When it comes to going from cc60 to cc30, it can make a big difference if you go from simple cc60 to simple cc30. It can't make a huge difference from double cc60 to simple cc30? There is just as much wood in the latter option. I'm asking because I have old beams that are uneven, etc., as described in a previous post. So, maybe it makes more of a difference how you apply reinforcement?huggan said:
I can't use self-leveling compound, but it's a good method because the compound on top doesn't compress easily, which is why the floor becomes stiff.
I could screw metal strips on the bottom or the underside of each new beam I install to achieve tensile strength at the bottom, to prevent expansion as much as possible. The question is just how much it does, but, of course, it depends on the dimension of the steel strip.
Further thoughts I've had are to take a 45x120 and glue-nail a horizontal 45x95 or 45x120 on the bottom to make it like an upside-down T, or to glue-nail a 45x45 on each side at the bottom of a 45x170, so that also becomes like an upside-down T.
Both of these options should provide additional reinforcement. But considering that you might not use the best adhesive for the purpose and don't glue under optimal conditions, perhaps the latter option is recommended over the first, so the height of 170 mm consists of one single piece of wood, not to mention that it's 5 mm more than 120+45 mm. Otherwise, reinforcement adheres significantly better according to the first option technically.
Clearly, this kind of thing requires thorough calculations/tests to be able to provide credible answers
Yes, it's the same amount of wood, yes it can withstand the same load. But the stiffness won't be as good as cc 300.Martin_B said:
Ok. Yes, as mentioned, it feels better distributed at cc30, definitely. However, I'm still somewhat curious about the physics behind it. Do you know what makes it stiffer? Because at the same time, you have to consider that there will be no glue and nails (however nails affect it). But the glue should have a positive effect?huggan said: