Parquet Floor: Expansion Gap???

[pk00213]

I would appreciate help with a query concerning the need or otherwise for a perimeter expansion gap with parquet flooring.

I intend laying parquet (280x70x10 without T/G) flooring in my hall. It's a suspended floor. The plan is:

Remove any highpoints on existing floorboards and ensure all firmly secured to joists.
Overlay with 9mm ply, screwed to floorboards.
Fix parquet with appropriate adhesive (not yet chosen it).

I would prefer to scribe parquet to items such as the bottom of staircase which has curved profile, several threshholds, etc and not leave an expansion gap. This would produce a far neater job than leaving gaps to be filled with (unsightly) cork or trying to mask gaps with scotia. But I've read various posts re flooring and many refer to a gap being an absolute must. I've got a problem trying to understand the logic of this advice.

If the new floor is floating (e.g. laminate flooring planks) then yes, I can see that you definitely need provison for movement at the perimeter because the floor is free to expand/contract relative to its surroundings. However, if the new floor (parquet), the ply sub-floor, the original floor boards and the joists are all rigidly secured together then relative movement cannot occur i.e. the different layers will act compositely by virtue of the nails, screws, and adhesive at the respective interfaces. Thus the only way that large movements can accumulate at the perimeter is if the the whole caboodle moves (including the joists) relative to the walls of the house. Is this what those who advocate a perimeter expansion gap believe really happens???

 

[WoodYouLike]

Is this what those who advocate a perimeter expansion gap believe really happens???

We don't only believe it - we've seen it happening.
Wood, no matter how thin or narrow, works due to changes in humidity. If you can make sure the humidity in your home is always constant - which you can't - then you are right.

In all other circumstances: LEAVE AN EXPANSION GAP OF AT LEAST 10MM ALL AROUND

 

[pk00213]

Thanks for taking the trouble WoodYouLike. Much appreciated.

We don't only believe it - we've seen it happening.

I'm sure you have seen many examples of distressed floors due to lack of adequate provision for movement but surely they must have occurred because either

1. The complete floor moved i.e. parquet, ply, floorboards and joists acted compositely. This is very unlikely as it is not feasible that 10mm thick parquet could 'drive' 18mm ply plus 21mm oak floorboards plus heavy joists timbers.

2. The restraint between the parquet and ply (adhesive and/or nails) has failed.

3. Failure has occurred within the fibres of the wood itself (i.e. the adhesive perfoms but the timber delaminates).

In your experience, which of these scenarios occurs?

Wood, no matter how thin or narrow, works due to changes in humidity.

I understand what you are getting at. Timber wants to change its volume dependent on its moisture content but it can only do so if it is not restrained. To take an extreme example, if you bolted a 10mm thick sheet of timber to a 25mm steel plate such that it was fully restrained in both lateral directions, then it would be unable to expand/contract in the plane of the timber. If humidity rose above the value that applied at the time of bolting, then instead of expanding the timber would go into compression. If humidity dropped below the value at the time of bolting, instead of contracting the timber would go into tension.

So to apply that principle to the parquet adhering to a plwood sub-floor, IF adhesives can be relied on to fully restrain the parquet with respect to the ply in both lateral directions, then relative movement cannot occur. Likewise if the ply/floorboard/joist act compositely then the perimeter 'expansion gap' would not be necessary.

Are the failures you see down to inadequate adhesive performance?

Thanks in anticipation.

 

[WoodYouLike]

No, these filling of the gaps all around the perimeter of the floor (10mm around the 10mm blocks and even the 10mm around the plywood sheets) are down to the wood (blocks and plywood) moving due to changes in the air humidity.

You don't have to believe me - we're only installing wooden floors, and especially design parquet floors, for around 14 years, so what do we know? - and test out your own thoughts.

 

[nickdsjnk]

Last year I finished installing some teak (or similar) parquet in my house. I laid 18mm plywood over the pitch pine floorboards (on joists), and bonded the parquet to the plywood with lecol 5500. I left an expansion gap absolutely everywhere, including door frames, the stair riser, doorways, skirting boards, etc. I left about 12mm

The tiles are small (about 23mm x 115mm) and the rooms are no longer than 5m.

I measure the gap every few months in a couple of small place where it is not covered and I haven't measured a single mm of movement. That is with humidity ranging from 40% to 80% (according to my dehumdifier) and temperatures from 13deg to 25deg.

There is really no way any of it could move, but I followed the expansion gap advice anyway as the extra work was a small price to pay for the possibility of wasting so much time and materials. The joists are tied into the brickwork, the floorboards are nailed to the joists, the plywood is screwed through the floorboards to the joists and the parquet is bonded on top. If the parquet moved, the whole house would be squeeking and cracking.

 

[AronSearle]

Thanks for taking the trouble WoodYouLike. Much appreciated.

We don't only believe it - we've seen it happening.

I'm sure you have seen many examples of distressed floors due to lack of adequate provision for movement but surely they must have occurred because either

1. The complete floor moved i.e. parquet, ply, floorboards and joists acted compositely. This is very unlikely as it is not feasible that 10mm thick parquet could 'drive' 18mm ply plus 21mm oak floorboards plus heavy joists timbers.

2. The restraint between the parquet and ply (adhesive and/or nails) has failed.

3. Failure has occurred within the fibres of the wood itself (i.e. the adhesive perfoms but the timber delaminates).

In your experience, which of these scenarios occurs?

Wood, no matter how thin or narrow, works due to changes in humidity.

I understand what you are getting at. Timber wants to change its volume dependent on its moisture content but it can only do so if it is not restrained. To take an extreme example, if you bolted a 10mm thick sheet of timber to a 25mm steel plate such that it was fully restrained in both lateral directions, then it would be unable to expand/contract in the plane of the timber. If humidity rose above the value that applied at the time of bolting, then instead of expanding the timber would go into compression. If humidity dropped below the value at the time of bolting, instead of contracting the timber would go into tension.

So to apply that principle to the parquet adhering to a plwood sub-floor, IF adhesives can be relied on to fully restrain the parquet with respect to the ply in both lateral directions, then relative movement cannot occur. Likewise if the ply/floorboard/joist act compositely then the perimeter 'expansion gap' would not be necessary.

Are the failures you see down to inadequate adhesive performance?

Thanks in anticipation.

What you are saying is somewhat sound in theory, you are essentially making a giant sheet of plywood, but...

1. plywood still expands, less than solid timber, but enough that you still need a gap. That gap might be 2-5mm instead of 10mm, not sure it's worth it myself.

2. Find an adhesive manufacturer that will guarantee a site applied bond that will perform as you envision (in writing).


I have seen factory made engineered boards ripple or buckle (the boards themselves not the floor) as the movement was greater than allowed for coupled with adhesive failure (and no, they were not made in china!), you want to do on site, what some people have failed to do in factories, good luck.



Seeing as it is well documented that a timber floor can push out masonary walls, if no expansion room is left, you really need to take a cost vs risk approach.

Personally wouldst consider doing it.

 

[WoodYouLike]

Are the failures you see down to inadequate adhesive performance?

Modern adhesives (flexible ones) will accept a certain level of movement of the wood without breaking their bond.

However, excessive movement of the wood blocks themselves - due to changes in humidity - will not be stopped by the adhesive itself (or by any other installation method). We cannot change the laws of physics I'm afraid.

 

[pk00213]

Thanks nickdsjnk

There is really no way any of it could move, but I followed the expansion gap advice anyway as the extra work was a small price to pay for the possibility of wasting so much time and materials. The joists are tied into the brickwork, the floorboards are nailed to the joists, the plywood is screwed through the floorboards to the joists and the parquet is bonded on top. If the parquet moved, the whole house would be squeeking and cracking.

I agree entirely! I find your response reassuring as you have set down actual evidence from your own 'experiment' with actual measurement data. Thank you very much, it is most helpful.

Thanks AronSearle. Perhaps you're right and I should go for belt and braces, but I don't know..... Surely the movement can be engineered out by providing adequate restraint?

Thanks WoodYouLike but I feel that you may have overlooked the possibilty that timber movement can be restrained. There are adhesives which are stronger than timber in both shear and tension. nickdsjnk's floor is 'breathing' with changes in RH. As you correctly state "the laws of physics" dictate that the parquet would like to move but the evidence from his measurements at different RHs is that it is restrained from doing so. Instead his parquet must be experiencing compression/tension and the gaps he provided are redundant.

 

[WoodYouLike]

You're right, the wood is restrained - in most normal circumstances - by the flexible adhesive.
But.......
the minute the air humidity tumbles down sharply or increases tremendously over a short period of time (imagine a whole week of rain and overclouded sky, and then the sun comes out to evaporate all the moist) your wooden floor will shrink or expand, with the latter the flexible adhesive will keep it down but will not restrict the wood of expanding and therefore the need of expansion gaps.

 

[mac391]

pkoo213
Either listen to the advice you have asked for from experienced floor layers, which i read as 100% correct or...

Conduct your own experiment