I've asked loads of advice regarding my mouldy kitchen floor.
Thought some may be interested in an update and some techniques I've been
using...
The kitchen (former living room) is about 20m2, old garage knock-through,
floor more than an inch out of true, mostly in small concentrated areas
which meant you could feel it just by walking. Chances of laying tiles on
that (unless they were mosaic) - about zero...
1) Sub job out to builders with simple specification of "please flatten" and
the caveat "is that black tile adhesive residue likely to be a
problem". "Sure" they said "and no, latex screed sticks to everything"...
2)
Get this:
http://photos.dionic.net/v/public/bu..._0002.jpg.html
http://photos.dionic.net/v/public/bu..._0003.jpg.html
Oh dear - still not flat (never mind level) and 3 months later, several
sections were becoming loose. Started investigations on uk.d-i-y and
t'internet.
3) "Develop" a simple height surveying technique based on a Stanley
cross-beam self levelling laser:
http://photos.dionic.net/v/public/bu..._0003.jpg.html
4) Find out that black stuff was "Blackjack" to which nothing sticks, and
cretinous builders used cement/silica gunk and not even latex gunk.
5) SDS up crap and run tungsten tipped flail drum grinder + industrical
vacuum over floor. Based on survey in 3, manage to grind one bay section
down 1cm which brings the whole thing to around 15mm total variation rather
than the 25mm before. Bugger all blackjack left except for odd very small
patches.
Here's the result, after infinite hoovering:
http://photos.dionic.net/v/public/bu..._0024.jpg.html
6) Cut out detatched screed sections - 2m2 by door and couple of other
little bits (sq foot each). Screed is only 2cm thick at best.
7) Notice floor is *incredibly* damp. Notice that 1/2 room (garage bit) has
excellen (if lumpy) screed and other half is made from 99% plastering sand.
8) Soak whole room in 10l SBR + 30l water. Keep going until screed sucks no
more.
9) Get this:
http://photos.dionic.net/v/public/bu..._0001.jpg.html
The brown screed is the crap - but even this is now solid. Cut a bit out and
verify it is solid all the way through. Isn't SBR wonderful!
10) Do a new survey and get this result:
http://photos.dionic.net/v/public/bu...ZPlan.png.html
"0" is the nominal height of the hall floor. +2 means 2mm higher than
hall, -14 means 14mm lower - you get the idea.
11) Write perl program to take this data, and using the volume under a 3D
triangle formula, calculate the volumes of gunk required to take each
vertical strip to a new height - I decided on 2mm above hall floor as this
meant I didn't have to do any more grinding (ignore the +6 - I dealt with
this after that survey).
12) Decide to lay F Ball Stopgap 900 DPM Prep followed by F75 Epoxy DPM,
them primed with P131 and topped with Stopgap 300 based on excellent advice
from here, previous experience with F Ball products and some chats with F
Ball tech. Stopgap 900 is an acrylic compound and withstands damp. It's
primarily designed to act as a smoothing substrate for F75. Unfortunately
F75 must be topped off with something else, and Stopap 300 is a listed
option, so so be it.
13) Stopgap 900 is limited to 15mm (20mm with granite chipping filler).
Infill deep holes with SBR sand cement screed (10-15mm). Mix half bag of
Stopgap 900 + filler and fill small awkward bits - we figure we can bend
the rules a bit...
14) Find out we need 13-14 bags of Stopgap 900. With a working time of 20-30
minutes on a cool day, we are not going to be able to do that in one hit.
Design railing system like this:
http://photos.dionic.net/v/public/bu..._0001.jpg.html
http://photos.dionic.net/v/public/bu..._0004.jpg.html
http://photos.dionic.net/v/public/bu..._0007.jpg.html
http://photos.dionic.net/v/public/bu..._0014.jpg.html
That's ali channel, nice and stiff bolted down to brass studs (cut off
bolts) resin mortared into small 5-10mm deep holes in the floor. There's a
nut (or packing washers) under than channel to set the bottom of the
channel to the required finish height - all down with a 16mm oval conduit
sighting stick and a laser. Masking tape floor and channel, drop channel
on, bolt down, double check with 6' spirit level (all perfect) and floam
seal so gunk cannot escape. Masking tape makes for an easier cleanup of
rail and floor afterwards.
This stage takes about 2 hours per pair of rails (which are butted from
shorter sections for convenience).
The scraper bar is made from the same channel + 2 "ears" of smaller (could
be the same) stuff. The scraper bar inherently scrapes level with the
bottom of the main rails.
15)
Having done a pouring plan based on strips consuming no more than an
estimated 2.5 bags (which always means upto 3 bags due to errors and
wastage), mix a 3 bag batch (with a friend) and pour in. Quick push around
with a float to wet everything and pour some more until 70% is down.
Use scraper bar to pull excess to low areas.
The scraper bar can also be dropped on at random places to spot check the
level - if it *just touches* the gunk, all is good.
Pour more until we have filled the rails evenly. Run over with a spiked
roller (optional) to kick air bubbles out and recheck level with scraper.
Leave for 2 hours and carefull remove rails for reuse.
In theory Stopgap 900 has a 90 minute walk on time. I found it to be rather
longer - it's solid after that time, but not walkable. Probably because my
subfloor is more water resistant due to the SBR than the more usual case
of pouring it directly onto concrete. I bent the rules here, but the first
strip:
http://photos.dionic.net/v/public/bu..._0015.jpg.html
http://photos.dionic.net/v/public/bu..._0017.jpg.html
was as hard as iron the next day, and very well adhered to the SBR so I
judge no ill effects. F Ball won't guarantee this combination but I'm happy
which what I've seen. It's either that or pouring it onto friable crap
which seems a less good idea. If we dig out the firable crap we have vastly
more filling to do and just have to bind it to friable concrete instead. So
we'll call this an "engineering solution".
Lose excess mix in very low spots like this:
http://photos.dionic.net/v/public/bu..._0018.jpg.html
All in all, it took me 1 day to do that strip including cutting the ali,
planning and trying stuff.
Me and a mate did two deep strips yesterday - a lot of the time wasted was
waiting for either PU foam to dry or gunk to be solid enough to go near.
I now have strips A, B and C poured and fingers crossed, they will be level
and mutually at the same height. A certainly checked out to be within 1mm
in any direction.
Now I just have to mix 1 and 2 bag batches and infill, using the current
strips as my guides - so no more rail setup.
Was it worth it?
So far, it looks so. The rails were necessary - Stopgap 900 is less free
flowing than Stopgap 300 so it needed help. Being able to spot check along
the length was invaluable. Knowing I have 3 reference strips that are
correct and flat should lead to a floor with no more than a mm or two error
and that is more than acceptable.
I'll post some more at future stages too.
Cheers
Tim
--
Tim Watts
The ****artist formerly known as Tim S