ground floor dead load

[mcmook]

i'm planning to install UFH in a ground floor on a timber floor. the floor is not exactly suspended, more the joists are laid resting on bricks and stones at intervals over a solum which has had tar (mastic asphalt i think the BC guy said) poured as a DPC. the joist are in good nick so the DPC seems to have been doing its job.

i've been googling for dead load tables and they all seem to work the opposite way from which i need ie they quote permissable span for for a range of dead load. i need to know will my floor (approx 450mm centres 7"x3" joists) take a dead load of 30kg / m2 (which is the weight of a 25mm screed required for the UFH).

the joists run parallel with the longest wall and the area is about 70 m2 (which will require about 2 tonnes of screed)

could the spot loading ie where a stone/brick supports the joist cause any problems?

i'll ask an architect friend ... but i thought i'd try here too... advice is usually excellent

[The Natural Philosopher]

To be honest, I would take the whole floor up, put in insulation, and a fresh DPC, re lay the screed with the UFH and put the wood back over.

Top quality insulation is the order of the day - 75mm polystyrene at least..and at least 75mm of screed IMHO. You can brace the while thing with reinforcing mesh.

I think it owould be less work overall frankly.

And guaranteed quality once done.

[mcmook]

The floorboards are up already - Why the need for another DPC? As i said the joists are in good condition. So I was going to insulate between the joists (75mm) then lay a subfloor, 9mm OSB, then

25mm spacers on the joists with the pipework between and a 25mm screed (this is whats been recommended) by the UFH company, then the final floor covering.

What I'd like to know is wheter my joists will take the weight of the screed, and whether there may be any problems from spot loading where they are resting on the bricks/stones.

[mcmook]

Do you mean remove the joists and replace with a solid floor?

[John Rumm]

You don't mention the one vital bit of information - the span between supports for the joists. It also sounds from your description like you may have more than just one support either end of each joist? If this is the case then that adds considerable resistance to bending and reduces the maximum shear load on any one part of the joist.

However gut feel suggests it ought to be ok since A modern floor would probably have a design load of about 100kg/m^2, older ones may be less stiff, but not massively so. A modern floor would not usually have the benefit of any extra piers for midway support on the joists either. The screed itself will also resist bending.

[mcmook]

Thanks John... the one bit of vital information is a random variable! each joist is probably supported on average every 2-3m by a brick or stone, but i'll have a proper look this evening.

[John Rumm]

In which case that sounds pretty good.

If you take the worst possible case as a 3m span, and ignore the fact that you will be getting extra strength from the cantilevered sections of joist the other side of the piers, and yet more strength from the screed itself. You have a basis for an overly pessimistic calculation.

I did a quick calculation in SuperBeam with a joist as described, and upped the (typical) linear distributed load from 0.8kN/m to 1.2kN/m to take account (and then some) of the extra screed and the wider than normal joist spacing. Even with this extra load, the deflection in the centre of the joist is only about half that permitted by building regs today. Note also that building regs are specified so as to prevent damage to ceilings etc from sagging joists - there is significant extra strength available if you can tolerate the extra deflection. So in summary, it sounds ok to me (insert normal disclaimer here!)

[fred]

I think that NPs solution is a tad over the top, John has the joist loading experience and says it is ok so I would go with that, my only other concern was potential sinking of the piers (, half bricks or whatever) but a quick calc from your 2ton total, 70m2 area and 3m spans gives an extra loading of 40kg per pier or joist so little to be concerned about there.

Bottom line, go for it.

[The Natural Philosopher]

Oh why not?

Cheap as anything and more reliable to use a plastic sheet than a nbit of tar.

Well they might, but why take the risk? Get rid of all underfloor wood and fill the gap with insulation and screed...

Its more a question of screed cracking than joists being under spec..they will take the weight..but not without flexure. Flexing screed will crack. Period.

>

[The Natural Philosopher]

Yes.

[The Natural Philosopher]

I don't think it is.

Consider. You have done all the calcs, laid the screed, and your X year old kid jumps up and down in a two weeks later you have a cracked and crumbling screed.

For what? taking out the joists, laying insulation (which will be easier anyway) and screed (which you have to anyway) ..

The opportunity cost seems to me to be simply removing the joists..a work of less than a day surely.

The actual installation of the screed and insulation should be easier as a result, the DPC can be replaced with something that is modern and reliable - which bitumen is IMHO not very..

In short, for very little effort extra, probably less cost and certainly far more quality can be built in for no extra cost .. and its a

*guaranteed* result..its something that's been done time and again..

Its that an 'excessive' solution?

[mcmook]

hmm. these things tend to snowball. it all started because i had a (small) section of floor lifted to inspect the joists...may as well lift the floor and lay a subfloor cos sections of the floor were different board thicknesses...and if i do that i can insulate... if i insulate i may as well go for UFH...

if i laid insulation and a screed as you suggested i suspect i would need to do this on top of a concrete slab? i need to confirm this? if so i'd need 100mm concrete (=A3600) + insulation (=A3250) + reinforcing mesh (=A3150) - thats just materials - labour on top of that. i think it would be more work and expense but possibly a more robust solution as you've pointed out.

[fred]

I stick with my earlier analysis . . .

[The Natural Philosopher]

Those numbers seem very high to me..just what are of floor are you talking about? From memory floor grade polystyrene is about £4 a sq meter..are you really telling me you have 60 sq meters? Thats not a room, that's a whole house!

Are you saying the joists are just resting on earth?

[mcmook]

.=2E. in my first post i quote the floor area its 70 m2.

[The Natural Philosopher]

Ah..long time ago that post.

Mmm. Well you know the issues..its your cost benefit calculation now ;-)

I will just say that every time I have tried to 'adapt existing' in any kind of work, from fixing cars to fitting out houses..the cheaper way has generally been..

"WHEN IN DOUBT, RIP IT OUT"

Once you go beyond a simple repair, its usually best to cut back to sound, and use new..

You don't necessarily need rebar.

If your soil is stable you generally whack in hardcore, top up with 3-4" of concrete to make the slab then chuck in 3-4" of poly, a DPM and top fill with screed and water pipe.

The only extra expense I can see is the concrete and DPM.

Otherwise you were going to do the insulation and screed anyway...?

[Cicero]

=============================== It would also be sensible to review the *weight* of the screed required before a final decision is made. The OP's original post proposed a screed depth of 25mm. This seems very little when 50 / 75mm is considered the norm.

Cic.

[mcmook]

50 / 75 mm is the norm in a solid floor. In a suspended timber floor application 25mm is recommended this is a half way house between utilising some thermal mass (as opposed to using reflecting fins - no thermal mass / good response time) and using a full depth screed good thermal store / poor response time. This was recommended by my installer a reseller for Begatube.