I have a hole in a wooden floor as shown on the picture below.

enter image description here

There are different bumps on the underfloor beneath. There used to be and old furnace installed here before so it's possible that the underfloor has been additionally reinforced.

I've managed to find some leftover wood lying around that is thick enough to fill the hole to the level which the primary floor boards take.

enter image description here

However, at some places the holes in the underfloor are to big, so the the new boards would either bend or float.

Here is a conceptual drawing of the situation:

enter image description here

1. This is how it currently looks.

2. and 3. are the approaches I would choose to take.

Is there a established best practice on how one would approach this problem and what could be the pitfalls?

Making a stable support for the frame is turning into more work that I have anticipated. Some of the boards of the underfloor are having almost 1cm hight differences.

So far, the simples way I found to stabilize the boards was to use wedges and flatten the pointing edge so it serves as a touch point for the neighbor board.

enter image description here enter image description here

  • A couple questions: 1) What's the stuff that looks like mortar or something between the subfloor boards? 2) Is the subfloor in that location laying on a masonry foundation, perhaps to support the weight of the furnace? 3) Of what material are the "high-spots"? Is it wood that could be surfaced down or masonry that would need chipping away with a cold chisel?
    – scanny
    Mar 2, 2017 at 1:03
  • @scanny 1) Leftovers from removing the furnace. Unfortunately I wasn't there when it was being taken away. 2) Don't know what's underneath, it's a hole when I try with a screwdriver, doesn't seem like there is masonry close below. 3) The high spots are wooden underfloor levels. Mar 2, 2017 at 1:15
  • 1
    How thick is the top-level flooring? It looks like tongue-and-groove, which in my house is 1-1/2" thick. If that's the case it will have no problem spanning say 24" at least. You'd just need to scrap together solid support points by either belt-sanding or building up as appropriate (maybe with different thicknesses of framing lumber ripped to the appropriate thickness). Wood that thick would span the low spots (unless you want a "creak-free" floor :) You might want to see if you could rout a tongue and groove or two grooves and spline for the ends where they join since they're short pieces.
    – scanny
    Mar 2, 2017 at 7:14
  • @scanny I'm currently testing various methods of building the framing lumber, see my answer. Don't understand what you mean by Wood that thick would span the low spots (unless you want a "creak-free" floor. Mar 3, 2017 at 20:44
  • Creaking (often "squeaking" on Google it appears) happens when adjacent boards rub against each other or move in some way under load (foot traffic). A "squeak-free" floor is one where the boards (or more likely panels) are fixed (perhaps glued) in such a way to absolutely prevent movement. I don't think you're going to achieve that here. Thick boards will span (like not break) the high spots, but may still deflect enough to produce noise. Google has a lot of resources on it, this is one: newhomesource.com/resourcecenter/articles/…
    – scanny
    Mar 3, 2017 at 21:48

2 Answers 2


Your option 2 is, in essence, how many uneven subfloors are accounted for. In construction and floor laying it's quite common to use shims or packing to level out irregularities and then to fix through them to the material on the other side/below.

The exaggeration of the 'humpy' nature of the floor in the drawing may be misleading as to the severity of the problem so I think this approach should work, particularly as it's a corner which presumably will not be a high-traffic area.

However, you may way to take care that the edges of every board are supported so that they don't tend to spring when stepped on.

Grain orientation
The new piece in photo no. 2, if it's solid wood, has grain oriented such that expansion and contraction will be along its length and not across its width the way it is in floorboards normally. This could present a problem as at maximum expansion (during the wettest time of the year) it will lengthen significantly. It's possibly at maximum contraction now since it's winter, so the difference between the current 'length' of a piece such as this and how it'll be later in the year could be significant, easily over 1cm.

  • has grain oriented such that expansion and contraction will be along its length and not across its width the way it is in floorboards normally could you please extend the answer with how one can know this by looking at the wood piece? Mar 2, 2017 at 10:20
  • 1
    @TheMeaningfulEngineer Wood expands across the grain, not along it, so usually boards change width but not length because the grain runs along the long dimension. Here the grain is at 90° to the long dimension, so it will change in length (assuming it's solid wood as it appears to be from the top surface). See previous Questions here on grain orientation and wood movement if you want more info.
    – Graphus
    Mar 2, 2017 at 14:44

This is not woodworking per se, but.... I would recommend mixing mortar and creating a level surface at one level in which the subfloor can be firmly supported and either anchored or glued. If the base concrete is loose, it should be removed down to a firm base and then refilled to the appropriate level. Note that anchoring to the concrete may be difficult with nails and may require concrete screws in drilled openings to provide a secure base.

The pitfalls are instability from not providing a firm anchorage to support the flooring and the possibility of mold or decay if the wood comes in contact with water or excessive humidity below.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.